Search

• A computational frame was proposed to predict diabetes that collected from hospitals and health center by leveraging blood routine data. • The contributions of different blood routine indicators for diabetes were identified by our frame. • A nomogram was constructed for assessing the influence of blood routine indicators on prediction outcomes. Diabetes stands as one of the most prevalent chronic diseases globally. The conventional methods for diagnosing diabetes are frequently overlooked until individuals manifest noticeable symptoms of the condition. This study aimed to address this gap by collecting comprehensive datasets, including 1000 instances of blood routine data from diabetes patients and an equivalent dataset from healthy individuals. To differentiate diabetes patients from their healthy counterparts, a computational framework was established, encompassing eXtreme Gradient Boosting (XGBoost), random forest, support vector machine, and elastic net algorithms. Notably, the XGBoost model emerged as the most effective, exhibiting superior predictive results with an area under the receiver operating characteristic curve (AUC) of 99.90% in the training set and 98.51% in the testing set. Moreover, the model showcased commendable performance during external validation, achieving an overall accuracy of 81.54%. The probability generated by the model serves as a risk score for diabetes susceptibility. Further interpretability was achieved through the utilization of the Shapley additive explanations (SHAP) algorithm, identifying pivotal indicators such as mean corpuscular hemoglobin concentration (MCHC), lymphocyte ratio (LY%), standard deviation of red blood cell distribution width (RDW-SD), and mean corpuscular hemoglobin (MCH). This enhances our understanding of the predictive mechanisms underlying diabetes. To facilitate the application in clinical and real-life settings, a nomogram was created based on the logistic regression algorithm, which can provide a preliminary assessment of the likelihood of an individual having diabetes. Overall, this research contributes valuable insights into the predictive modeling of diabetes, offering potential applications in clinical practice for more effective and timely diagnoses. [ABSTRACT FROM AUTHOR]

• Identification of DHSs can help to understanding the mechanism of disease development and the treatment of the disease. • Multi-dimensional feature fusion strategy was used for feature extraction from DNase samples. • An overall prediction accuracy of 0.859 was achieved with an AUC value of 0.837. DNase I hypersensitive sites (DHSs) are chromatin regions highly sensitive to DNase I enzymes. Studying DHSs is crucial for understanding complex transcriptional regulation mechanisms and localizing cis -regulatory elements (CREs). Numerous studies have indicated that disease-related loci are often enriched in DHSs regions, underscoring the importance of identifying DHSs. Although wet experiments exist for DHSs identification, they are often labor-intensive. Therefore, there is a strong need to develop computational methods for this purpose. In this study, we used experimental data to construct a benchmark dataset. Seven feature extraction methods were employed to capture information about human DHSs. The F -score was applied to filter the features. By comparing the prediction performance of various classification algorithms through five-fold cross-validation, random forest was proposed to perform the final model construction. The model could produce an overall prediction accuracy of 0.859 with an AUC value of 0.837. We hope that this model can assist scholars conducting DNase research in identifying these sites. [ABSTRACT FROM AUTHOR]

Risk stratification for poor outcomes is not currently age-specific. Risk stratification of older patients based on observational cohorts primarily composed of young patients may result in suboptimal clinical care and inaccurate quality benchmarking. We assessed two hypotheses. First, we hypothesized that risk factors for poor outcomes after trauma are age-dependent and, second, that the relative importance of various risk factors are also age-dependent. A cohort study of severely injured adult trauma patients admitted to the intensive care unit 2014-2018 was performed using trauma registry data. Random forest algorithms predicting poor outcomes (death or complication) were built and validated using three cohorts: (1) patients of all ages, (2) younger patients, and (3) older patients. Older patients were defined as aged 55 y or more to maintain consistency with prior trauma literature. Complications assessed included acute renal failure, acute respiratory distress syndrome, cardiac arrest, unplanned intubation, unplanned intensive care unit admission, and unplanned return to the operating room, as defined by the trauma quality improvement program. Mean decrease in model accuracy (MDA), if each variable was removed and scaled to a Z-score, was calculated. MDA change ≥4 standard deviations between age cohorts was considered significant. Of 5489 patients, 25% were older. Poor outcomes occurred in 12% of younger and 33% of older patients. Head injury was the most important predictor of poor outcome in all cohorts. In the full cohort, age was the most important predictor of poor outcomes after head injury. Within age cohorts, the most important predictors of poor outcomes, after head injury, were surgery requirement in younger patients and arrival Glasgow Coma Scale in older patients. Compared to younger patients, head injury and arrival Glasgow Coma Scale had the greatest increase in importance for older patients, while systolic blood pressure had the greatest decrease in importance. Supervised machine learning identified differences in risk factors and their relative associations with poor outcomes based on age. Age-specific models may improve hospital benchmarking and identify quality improvement targets for older trauma patients. [ABSTRACT FROM AUTHOR]

Accelerometry and machine learning are powerful tools for gaining in-depth information about animal behaviour. However, such methods are rarely used to quantify behaviours that contribute to mating success, such as intrasexual competition, courtship and copulation. In part, this is probably due to the inherent challenges of classifying behaviours that can be brief and infrequent, and that are not necessarily exhibited by all individuals. In this study, we used triaxial accelerometry and machine-learning methods to quantify the mating behaviours of captive male ruffs, Calidris pugnax. The ruff is a polymorphic, lekking shorebird with highly skewed mating success. Mating behaviour in ruffs includes ritualized postures that can be distinguished during observations. Using this system as a case study, we (1) describe possible approaches to classifying mating behaviour; (2) compare the classification performance of three supervised machine-learning methods: random forests, hidden Markov models and neural networks; (3) highlight potential pitfalls that can cause overestimation of model performance; and (4) offer suggestions for avoiding these pitfalls. In our study, some models distinguished mating behaviours from nonmating behaviours with high precision and sensitivity (>75%), but only when trained and tested on the same individuals within the same timeframe. Estimates of model performance were much poorer when models were tested on future data or data from different individuals. Nevertheless, even when tested on different individuals, the hidden Markov model provided a reasonably accurate estimate of which males invested more time in mating behaviour overall. Here, we provide an end-to-end workflow for classifying behaviour from accelerometry, including recommendations, considerations and code. Although posture-based mating behaviours in ruffs proved challenging to distinguish, the methods that we describe show promise for displays associated with distinctive, dynamic movements. • Accelerometry and machine learning are powerful tools for quantifying behaviour. • 'Mating' behaviours, such as courtship, can be challenging to classify reliably. • We describe and compare methods, using a lekking shorebird as a case study. • Models were used to rank males by time spent on mating behaviours with high accuracy. • Performance is overestimated when data are split randomly for training and testing. [ABSTRACT FROM AUTHOR]

A complex field environment and differences in maize density cause feed rate fluctuations in the combine. Traditional threshing methods do not adjust multiple operating parameters based on the feed rate, resulting in low threshing performance and high power consumption. A multi-parameter control method for maize threshing based on the feed rate was proposed to overcome these problems. Random forest, support vector machine, and multiple linear regression machine learning algorithms were used to estimate the threshing performance indices for different feed rates. The algorithm (random forest) was used to construct a control model for the rotor speed, threshing gap, and top cover guide vane angle. The performance of the control system was compared with a threshing system with constant parameters to verify the control model's accuracy. The results demonstrated that the threshing unit with the control system outperformed the one without. The broken grain rate and the power consumption were 2.18% and 36.22% lower, respectively, whereas the unthreshed grain rate was not significantly different. Random forest was more suitable than the support vector machine and multiple linear regression models for establishing a multi-parameter control model for maize threshing. The adjustment of multiple threshing parameters significantly improved the operating performance for different feed rates. • The threshing performance will decrease due to the fluctuation of feed rate. • Threshing multi-parameter control can effectively improve operation performance. • Artificial intelligence algorithm is conducive to the decision of harvest parameters. [ABSTRACT FROM AUTHOR]

This study explores the possibility of using routinely taken blood tests in the diagnosis and triage of patients with suspected musculoskeletal malignancy. A retrospective study was performed on results of patients who had presented for assessment to a regional musculoskeletal tumour unit. Blood results of patients with a histologically confirmed diagnosis between 2010 and 2020 were retrieved. 33 distinct blood tests were available for model forming. Results were standardised by calculating z-scores. Data were split into a training set (70%) and a test set (30%). The training set was balanced by resampling underrepresented classes. The random forest algorithm performed best and was selected for model forming. Receiver operating characteristic curves were used to find the optimum threshold. Models were calibrated and performance metrics evaluated with confusion tables. 2371 patients formed the study population. 1080 had a malignant diagnosis in one of three categories: sarcoma, metastasis, or haematological malignancy. 1291 had a benign condition. Metastasis could be predicted with an accuracy of 79% (AUC 87%, sensitivity 79%, specificity 80% NPV 91%). Haematological malignancy accuracy 79% (AUC 81%, sensitivity 77%, specificity 79%, NPV 97%). Sarcoma accuracy 64% (AUC 73%, sensitivity 76%, specificity 61%, NPV 88%) and all malignancy accuracy 74% (AUC 80%, sensitivity 72%, specificity 75%, NPV 76%). Routinely performed blood tests can be useful in triage of musculoskeletal tumours and can be used to predict presence of musculoskeletal malignancy. [ABSTRACT FROM AUTHOR]

The aim was to evaluate if maize (Zea mays L.) grain and silage yield can be estimated from unmanned aerial systems (UAS) imagery. A fixed-wing UAS was used to collect imagery throughout the growing season (emergence to reproductive growth stages) at eight commercial fields, each implemented with a nitrogen-rich (NRich) strip. Exponential regression models were fitted between yield and seven vegetation indices (VIs) from NRich strip, whilst two machine learning (ML) regression models (random forest [RF] and support vector regression [SVR]) were fitted with VIs, elevation, and landform layers. The results showed that the normalised difference vegetation index (NDVI) was the most reliable indicator for yield using exponential models when the images were collected at late vegetative stages before tasseling for silage and late reproductive stage before senescence for grain. The exponential models performed better with NRich strip than those derived with the statistical reference strips (SRS) for fields without NRich strips, but the differences were minimal when sensed at a vegetative stage before tasselling. The Boruta algorithm identified elevation, landform, and three VIs as the top five important features for yield estimation. Using these five features, the RF model produced a lower error and was faster to train than the SVR. The ML models outperformed the exponential models with SRS but underperformed compared to the models with NRich strip. It was concluded that if a field does not have an NRich strip, using RF models with additional data layers can provide reliable yield estimations for maize grain and silage. • Normalised difference vegetation index from UAS is a reliable indicator of yield. • Implementing NRich strip in the field helps in reliable yield prediction. • Sensing at V8–V10 growth stage was the best time for maize silage yield prediction. • Sensing at R4 was the best time for maize grain yield prediction. • Random forest model with five features can be used if NRich is not available. [ABSTRACT FROM AUTHOR]

[Display omitted] • CO 2 adsorption onto porous carbons was modelled by machine learning (ML). • Six ML models used, and feature importance and instance plots explained by SHAP. • GBDT showed the best prediction, however other models were also performed well. • CO 2 uptake more governed by adsorption conditions followed by adsorbent properties. Adsorption of CO 2 on porous carbons has been identified as one of the promising methods for carbon capture, which is essential for meeting the sustainable developmental goal (SDG) with respect to climate action, i.e., SDG 13. This research implemented six supervised machine learning (ML) models (gradient boosting decision tree (GBDT), extreme gradient boosting (XGB), light boost gradient machine (LBGM), random forest (RF), categorical boosting (Catboost), and adaptive boosting (Adaboost)) to understand and predict the CO 2 adsorption mechanism and adsorption uptake, respectively. The results recommended that the GBDT outperformed the remaining five ML models for CO 2 adsorption. However, XGB, LBGM, RF, and Catboost also represented the prediction in the acceptable range. The GBDT model indicated the accurate prediction of CO 2 uptake onto the porous carbons considering adsorbent properties and adsorption conditions as model input parameters. Next, two-factor partial dependence plots revealed a lucid explanation of how the combinations of two input features affect the model prediction. Furthermore, SHapley Additive exPlainations (SHAP), a novel explication approach based on ML models, were employed to understand and elucidate the CO 2 adsorption and model prediction. The SHAP explanations, implemented on the GBDT model, revealed the rigorous relationships among the input features and output variables based on the GBDT prediction. Additionally, SHAP provided clear-cut feature importance analysis and individual feature impact on the prediction. SHAP also explained two instances of CO 2 adsorption. Along with the data-driven insightful explanation of CO 2 adsorption onto porous carbons, this study also provides a promising method to predict the clear-cut performance of porous carbons for CO 2 adsorption without performing any experiments and open new avenues for researchers to implement this study in the field of adsorption because a lot of data is being generated. [ABSTRACT FROM AUTHOR]

Aflatoxin poisoning can harm broiler health and cause economic loss. Traditional diagnoses rely on human assessment, which is laborious and time-consuming. The objective of this study was to identify aflatoxin-poisoned broilers by analyzing three-dimensional accelerations collected by wearable accelerometers using five machine learning models, including K-Nearest Neighbor (KNN), Support Vector Classification (SVC), Decision Tree (DT), Random Forest (RF), and Gradient Boosting Decision Tree (GBDT). A total of 75 yellow feather broilers were reared in 3 pens (25 birds/pen) provided with 3 types of feed, including Normal (no Aflatoxin B 1 , control), AF50 (50 μg/kg Aflatoxin B 1 "AFB 1 "), and AF100 (100 μg/kg AFB 1). Lightweight tri-axial accelerometers were attached to 9 randomly selected birds (3 birds/pen) to continuously monitor the movement of broilers. Broiler behaviors were manually labelled as six categorises (sitting, standing, drinking, feeding, walking, and other) by observing the videos recorded by the camera installed in each pen. Different behavioral patterns were observed for AFB 1 poisoned birds as compared to that for sound birds. The poisoned broilers performed less standing, walking, feeding, and drinking behaviors, but more sitting behaviors. Five models could identify poisoned birds (AF100) with good accuracies (>86%). Higher misclassification rates were observed for all models in classifying AF50 and Normal groups. Among different models, the GBDT model performed best (97%) in differentiating Normal and AF50 broilers. In conclusion, three-dimensional accelerations and machine learning models could be used to identify aflatoxin-poisoned broilers with high accuracies. The findings could help in the development of viable systems that can identify diseased birds. • Poisoned broilers spent less time on feeding, drinking, walking, and standing. • Performances of models differ in identifying healthy and mild poisoning broilers. • The GBDT model performed better in identifying poisoned broilers. [ABSTRACT FROM AUTHOR]

Random feature methods have been successful in various machine learning tasks, are easy to compute, and come with theoretical accuracy bounds. They serve as an alternative approach to standard neural networks since they can represent similar function spaces without a costly training phase. However, for accuracy, random feature methods require more measurements than trainable parameters, limiting their use for data-scarce applications. We introduce the sparse random feature expansion to obtain parsimonious random feature models. We leverage ideas from compressive sensing to generate random feature expansions with theoretical guarantees even in the data-scarce setting. We provide generalization bounds for functions in a certain class depending on the number of samples and the distribution of features. By introducing sparse features, i.e. features with random sparse weights, we provide improved bounds for low order functions. We show that our method outperforms shallow networks in several scientific machine learning tasks. [ABSTRACT FROM AUTHOR]

The production of terpenoids from engineered microbes contributes markedly to the bioeconomy by providing essential medicines, sustainable materials, and renewable fuels. The mevalonate pathway leading to the synthesis of terpenoid precursors has been extensively targeted for engineering. Nevertheless, the importance of individual pathway enzymes to the overall pathway flux and final terpenoid yield is less known, especially enzymes that are thought to be non-rate-limiting. To investigate the individual contribution of the five non-rate-limiting enzymes in the mevalonate pathway, we created a combinatorial library of 243 Saccharomyces cerevisiae strains, each having an extra copy of the mevalonate pathway integrated into the genome and expressing the non-rate-limiting enzymes from a unique combination of promoters. High-throughput screening combined with machine learning algorithms revealed that the mevalonate kinase, Erg12p, stands out as the critical enzyme that influences product titer. ERG12 is ideally expressed from a medium-strength promoter which is the 'sweet spot' resulting in high product yield. Additionally, a platform strain was created by targeting the mevalonate pathway to both the cytosol and peroxisomes. The dual localization synergistically increased terpenoid production and implied that some mevalonate pathway intermediates, such as mevalonate, isopentyl pyrophosphate (IPP), and dimethylallyl pyrophosphate (DMAPP), are diffusible across peroxisome membranes. The platform strain resulted in 94-fold, 60-fold, and 35-fold improved titer of monoterpene geraniol, sesquiterpene α-humulene, and triterpene squalene, respectively. The terpenoid platform strain will serve as a chassis for producing any terpenoids and terpene derivatives. • A combinatorial yeast strain library tested the promoter space of MVA pathway enzymes. • Random forest modeling revealed the contribution of MVA enzymes to terpenoid production. • The mevalonate kinase, Erg12p, is critical for determining terpenoid yield. • Mevalonate and IPP/DMAPP are diffusible across peroxisome membranes in vivo. • A yeast platform strain is constructed as a chassis to produce any terpenoids. [ABSTRACT FROM AUTHOR]

Spectral analyses have become dependable and promising analytical methods for objective evaluation of coffee bean quality. However, although commercial instruments are available, financially accessible device solutions still need to be provided for this technique to become more widely available, particularly among emerging small and medium-scale coffee enterprises. This study developed an innovative multichannel spectral acquisition system integrated with an Internet of Things (IoT) platform for evaluating roasted coffee beans. The system was built using commercially available low-cost components with a total cost of approximately 114 USD. A study on the roasting degree classification combined with five machine learning algorithms was conducted based on 18 channels of spectral data (410–940 nm) acquired from the proposed device. Original spectral datasets were directly used for model development to minimize data processing and simplify the implementation of the best machine learning model on the device. The results revealed that the Random Forest (RF) model demonstrated a satisfactory performance (validation accuracy values reaching 0.988, precision 0.988, recall 0.988, and F1-score 0.987). Therefore, the proposed system can effectively classifiy roasted coffee beans and might have important applications in assisting roasteries during the roasting process in a real-time, non-subjective, and non-invasive manner. [Display omitted] • A low-cost IoT-based multichannel spectral acquisition device was developed. • A study on 18 spectral data was employed for four roast coffee beans degrees. • Roast coffee degree classification was performed using machine learning. • The best machine learning algorithm was found with a validation accuracy of 0.988. • This prototype has the potential to be used for broad agri-food product assessment. [ABSTRACT FROM AUTHOR]

The excessive application and loss of pesticides poses a great risk to the ecosystem, and the environmental safety assessment of pesticides is time-consuming and expensive using traditional animal toxicity tests. In this work, a pesticide acute toxicity dataset was created for silkworm integrating extensive experiments and various common pesticide formulations considering the sensitivity of silkworm to adverse environment, its economic value in China, and a gap in machine learning (ML) research on the toxicity prediction of this species, which addressed the previous limitation of only being able to predict toxicity classification without specific toxicity values. A new comprehensive voting model (CVR) was developed based on ML, combined with three regression algorithms, namely, Bayesian Ridge (BR), K Neighbors Regressor (KNN), Random Forest Regressor (RF) to accurately calculate lethal concentration 50 % (LC 50). Three conformal models were successfully constructed, marking the first combination of conformal models with confidence intervals to predict silkworm toxicity. Further, the mechanism by analyzing structural alerts was summarized, and identified 25 warning structures, 24 positive compounds and 14 negative compounds. Importantly, a novel comprehensive prediction system was constructed that can provide LC 50 and confidence intervals, structural alerts analysis, lipid-water partition coefficient (LogP) and similarity analysis, which can comprehensively evaluate the ecological toxicity risk of substances to make up for the incomplete toxicity data of new pesticides. The validity and generalization of the CVR model were verified by an external validation set. In addition, five new, low-toxic and green pesticide alternatives were designed through 50,000 cycles. Moreover, our software and ST Profiler can provide low-cost information access to accelerate environmental risk assessment, which can predict not only a single chemical, but also batches of chemicals, simply by inputting the SMILES / CAS / (Chinese / English) name of chemicals. [Display omitted] • A high-quality pesticide acute toxicity dataset for silkworm was created. • A new comprehensive voting model was developed combined with conformal models. • The previous limitation of only predicting toxicity classification was broken. • A comprehensive prediction system and software can evaluate the ecotoxicity risk. • Five low-toxic green pesticide alternatives were designed and screened. [ABSTRACT FROM AUTHOR]

Machine learning models are often viewed as black boxes in landslide susceptibility assessment, lacking an analysis of how input features predict outcomes. This makes it challenging to understand the mechanisms and key factors behind landslides. To enhance the interpretability of machine learning models in wide-area landslide susceptibility assessments, this study uses the Shapely method to explore the contributions of feature factors from local, global, and spatial perspectives. Landslide susceptibility assessments were conducted using random forest (RF), support vector machine (SVM), and eXtreme Gradient Boosting (XGBoost) models, focusing on the geologically complex Sichuan–Tibet region. Initially, the study revealed the contributions of specific key feature factors to landslides from a local perspective. It then examines the overall impact of interactions among feature factors on landslide occurrence globally. Finally, it unveils the spatial distribution patterns of the contributions of various feature factors to landslide occurrence. The analysis indicates the following: (1) The XGBoost model excels in landslide susceptibility assessment, achieving accuracy, precision, recall, F1-score, and AUC values of 0.7815, 0.7858, 0.7962, 0.7910, and 0.86, respectively; (2) The Shapely method identifies the leading factors for landslides in the Sichuan–Tibet region as Elevation (3000–4000 m), PGA (1–2 g), NDVI (<0.5), and distance to rivers (<3 km); (3) Using the Shapely method, the study explains the contributions, interaction mechanisms, and spatial distribution patterns of landslide susceptibility feature factors across local, global, and spatial perspectives. These findings offer new avenues and methods for the in-depth exploration and scientific prediction of landslide risks. • Utilizing spatial distribution patterns of landslide susceptibility in Sichuan and Tibet using SHAP values. • Analysis of the effect of evaluation factors on landslide susceptibility was analyzed. • Explored the Shapley method's feasibility for assessing landslide susceptibility. • Explained the process of employing the XGBoost model for evaluating landslide susceptibility. [ABSTRACT FROM AUTHOR]

This study investigated the impact of climate change on flood susceptibility in six South Asian countries Afghanistan, Bangladesh, Bhutan, Bharat (India), Nepal, and Pakistan—under two distinct Shared Socioeconomic Pathway (SSP) scenarios: SSP1-2.6 and SSP5-5.8, for 2041–2060 and 2081–2100. To predict flood susceptibility, we employed three artificial intelligence (AI) algorithms: the K-nearest neighbor (KNN), conditional inference random forest (CIRF), and regularized random forest (RRF). Predictions were based on data from 2452 historical flood events, alongside climatic variables measured over monthly, seasonal, and annual timeframes. The innovative aspect of this research is the emphasis on using climatic variables across these progressively condensed timeframes, specifically addressing eight precipitation factors. The performance evaluation, employing the area under the receiver operating characteristic curve (AUC) metric, identified the RRF model as the most accurate, with the highest AUC of 0.94 during the testing phase, followed by the CIRF (AUC = 0.91) and the KNN (AUC = 0.86). An analysis of variable importance highlighted the substantial role of certain climatic factors, namely precipitation in the warmest quarter, annual precipitation, and precipitation during the wettest month, in the modeling of flood susceptibility in South Asia. The resultant flood susceptibility maps demonstrated the influence of climate change scenarios on susceptibility classifications, signalling a dynamic landscape of flood-prone areas over time. The findings revealed variable trends under different climate change scenarios and periods, with marked differences in the percentage of areas classified as having high and very high flood susceptibility. Overall, this study advances our understanding of how climate change affects flood susceptibility in South Asia and offers an essential tool for assessing and managing flood risks in the region. • Machine learning algorithms were applied to predict flood susceptibility in South Asia. • The RRF model had the highest accuracy (AUC = 0.94) in flood susceptibility modeling. • CMIP6 data used to predict flood susceptibility under SSP1-2.6 and SSP5-8.5 scenarios. • Projected flood susceptibility maps for 2041–2060 and 2081–2100 periods under climate change impacts. [ABSTRACT FROM AUTHOR]

Soil salinization and sodification, the primary causes of land degradation and desertification in arid and semi-arid regions, demand effective monitoring for sustainable land management. This study explores the utility of partial least square (PLS) latent variables (LVs) derived from visible and near-infrared (Vis-NIR) spectroscopy, combined with remote sensing (RS) and auxiliary variables, to predict electrical conductivity (EC) and sodium absorption ratio (SAR) in northern Xinjiang, China. Using 90 soil samples from the Karamay district, machine learning models (Random Forest, Support Vector Regression, Cubist) were tested in four scenarios. Modeling results showed that RS and Land use alone were unreliable predictors, but the addition of topographic attributes significantly improved the prediction accuracy for both EC and SAR. The incorporation of PLS LVs derived from Vis-NIR spectroscopy led to the highest performance by the Random Forest model for EC (CCC = 0.83, R2 = 0.80, nRMSE = 0.48, RPD = 2.12) and SAR (CCC = 0.78, R2 = 0.74, nRMSE = 0.58, RPD = 2.25). The variable importance analysis identified PLS LVs, certain topographic attributes (e.g., valley depth, elevation, channel network base level, diffuse insolation), and specific RS data (i.e., polarization index of VV + VH) as the most influential predictors in the study area. This study affirms the efficiency of Vis-NIR data for digital soil mapping, offering a cost-effective solution. In conclusion, the integration of proximal soil sensing techniques and highly relevant topographic attributes with the RF model has the potential to yield a reliable spatial model for mapping soil EC and SAR. This integrated approach allows for the delineation of hazardous zones, which in turn enables the consideration of best management practices and contributes to the reduction of the risk of degradation in salt-affected and sodicity-affected soils. • The effectiveness of a variety of auxiliary variables was assessed in predicting EC and SAR. • PLS latent variables (LVs) derived from Vis-NIR led to the highest performance for predicting EC and SAR. • Among the various ML models, Random Forest (RF) demonstrated the highest efficiency. [ABSTRACT FROM AUTHOR]

Exploring the potential of new models for mapping groundwater quality presents a major challenge in water resource management, particularly in Kanchanaburi Province, Thailand, where groundwater faces contamination risks. This study aimed to explore the applicability of random forest (RF) and artificial neural networks (ANN) models to predict groundwater quality. Particularly, these two models were integrated into cross-validation (CV) and bootstrapping (B) techniques to build predictive models, including RF-CV, RF-B, ANN-CV, and ANN-B. Entropy groundwater quality index (EWQI) was converted to normalized EWQI which was then classified into five levels from very poor to very good. A total of twelve physicochemical parameters from 180 groundwater wells, including potassium, sodium, calcium, magnesium, chloride, sulfate, bicarbonate, nitrate, pH, electrical conductivity, total dissolved solids, and total hardness, were investigated to decipher groundwater quality in the eastern part of Kanchanaburi Province, Thailand. Our results indicated that groundwater quality in the study area was primarily polluted by calcium, magnesium, and bicarbonate and that the RF-CV model (RMSE = 0.06, R2 = 0.87, MAE = 0.04) outperformed the RF-B (RMSE = 0.07, R2 = 0.80, MAE = 0.04), ANN-CV (RMSE = 0.09, R2 = 0.70, MAE = 0.06), and ANN-B (RMSE = 0.10, R2 = 0.67, MAE = 0.06). Our findings highlight the superiority of the RF models over the ANN models based on the CV and B techniques. In addition, the role of groundwater parameters to the normalized EWQI in various machine learning models was found. The groundwater quality map created by the RF-CV model can be applied to orient groundwater use. [Display omitted] • Machine learning models were explored in groundwater quality assessment. • The model RF-CV outperformed RF-B, ANN-CV, and ANN-B. • The normalized EWQI was used to decipher groundwater quality zones. • The role of variables was different in machine learning models. [ABSTRACT FROM AUTHOR]

Despite advances, readmission and mortality rates for surgical patients with colon cancer remain high. Prediction models using regression techniques allows for risk stratification to aid periprocedural care. Technological advances have enabled large data to be analyzed using machine learning (ML) algorithms. A national database of colon cancer patients was selected to determine whether ML methods better predict outcomes following surgery compared to conventional methods. Surgical colon cancer patients were identified using the 2013 National Cancer Database (NCDB). The negative outcome was defined as a composite of 30-d unplanned readmission and 30- and 90-d mortality. ML models, including Random Forest and XGBoost, were built and compared with conventional logistic regression. For the accounting of unbalanced outcomes, a synthetic minority oversampling technique (SMOTE) was implemented and applied using XGBoost. Analysis included 528,060 patients. The negative outcome occurred in 11.6% of patients. Model building utilized 30 variables. The primary metric for model comparison was area under the curve (AUC). In comparison to logistic regression (AUC 0.730, 95% CI: 0.725-0.735), AUC's for ML algorithms ranged between 0.748 and 0.757, with the Random Forest model (AUC 0.757, 95% CI: 0.752-0.762) outperforming XGBoost (AUC 0.756, 95% CI: 0.751-0.761) and XGBoost using SMOTE data (AUC 0.748, 95% CI: 0.743-0.753). We show that a large registry of surgical colon cancer patients can be utilized to build ML models to improve outcome prediction with differential discriminative ability. These results reveal the potential of these methods to enhance risk prediction, leading to improved strategies to mitigate those risks. • Adverse outcomes remain common after surgery for colon cancer. • Logistic regression is a common tool for predictive modeling of surgical outcomes. • Machine learning methods may improve predictions compared to logistic regression. • Improved risk quantification, applied broadly, may greatly reduce morbidity and cost. [ABSTRACT FROM AUTHOR]

• Improved mapping of 2′-O-methylated (Nm) and pseudouridine (ψ) residues by RiboMethSeq. • Random Forest (RF) model including scores for neighboring positions gives the best results. • The model was trained using human rRNA datasets and successfully validated on other rRNAs. Analysis of epitranscriptomic RNA modifications by deep sequencing-based approaches brings an essential contribution to the general knowledge on their precise locations and relative stoichiometry in cellular RNAs. To reveal RNA modifications, several analytical approaches have been proposed, including antibody-driven enrichment, analysis of RT-signatures and specific chemical treatments. However, analysis and interpretation of these massive datasets, especially for low abundant cellular RNAs (e.g. mRNA and lncRNA) is not easy nor straightforward, since the insufficient specificity and selectivity are leading to massive false-positive and false-negative identifications. The main issue in the application of these methods relies on a subjective classification of potentially modified positions, mostly based on arbitrarily defined threshold values for different scores. Such approach using pre-defined scores' values was revealed to be appropriate for limited complexity datasets (for tRNA and/or rRNA analysis), but application to longer reference sequences requires much better classification algorithms. In this work we applied a machine learning algorithm (Random Forest, RF) to create a predictive model for analysis of 2′-O-methylated sites in RNA using RiboMethSeq datasets. Model's training was performed on a large collection of human rRNA datasets with well-known modification profiles and the performance of the prediction was assessed using experimentally defined profiles for other eukaryotic rRNAs (S. cerevisiae and A. thaliana). Application of this Random Forest prediction model for detection of other RNA modifications and to more complex datasets is discussed. [ABSTRACT FROM AUTHOR]

To evaluate the performance of the classic machine learning algorithms and the effectiveness of various features, the iterative algorithms (i.e., support vector machine (SVM), and least-squares SVM (LS-SVM)) and non-iterative algorithms (i.e., random forest (RF) and naive bayes (NB)) for six feature schemes were performed to classify the ECG recordings. The ECG recordings were initially filtered with a 0.1 Hz − 12 Hz band pass filter. Then 80 features, including 48 time domain, 18 frequency domain, 12 time–frequency and two principle component analysis (PCA) features, were extracted to construct six feature schemes. The RF, SVM, LS-SVM and NB were employed to discern a binary-classification task (i.e., normal and AF ECG recordings) and a tri- classification task (i.e., the normal, AF and ST change ECG recordings) for the six feature schemes. The results revealed that time domain, frequency domain features and PCA features can provide relatively reliable feature combinations to the RF and SVM. In addition, the RF yielded the highest F1-scores (0.8908 and 0.7535) for the binary-classification task and the tri-classification task than the SVM, LS-SVM and NB. [ABSTRACT FROM AUTHOR]

• The devices and sensors in blockchain-enabled IoT network produce a lot of data, so AI is applied as analytical tool to provide consistent results in decision-making. • Fog computing paradigm is used to decentralize cloud based centralized security mechanism, thus data analysis and security related concerns are handled at edge of networks. • A distributed IDS is designed using fog computing to detect DDoS attacks against memory pool in blockchain-enabled IoT network. • To evaluate the proposed detection system two well-known machine learning algorithms, random forest and XGBoost are used in distributed architecture. • An actual IoT based BoT-IoT dataset is used to analyze the performance of the model. As it contains various recent Botnet related attacks such as DoS, DDoS, theft. • Different evaluation metrics such as accuracy, detection rate, false alarm rate, and precision are used to thoroughly investigate performance of the proposed. The Internet of Things (IoT) is emerging as a new technology for the development of various critical applications. However, these applications are still working on centralized storage architecture and have various key challenges like privacy, security, and single point of failure. Recently, the blockchain technology has emerged as a backbone for the IoT-based application development. The blockchain can be leveraged to solve privacy, security, and single point of failure (third-part dependency) issues of IoT applications. The integration of blockchain with IoT can benefit both individual and society. However, 2017 Distributed Denial of Service (DDoS) attack on mining pool exposed the critical fault-lines among blockchain-enabled IoT network. Moreover, this application generates huge amount of data. Machine Learning (ML) gives complete autonomy in big data analysis, capabilities of decision making and therefore is used as an analytical tool. Thus, in order to address above challenges, this paper proposes a novel distributed Intrusion Detection System (IDS) using fog computing to detect DDoS attacks against mining pool in blockchain-enabled IoT Network. The performance is evaluated by training Random Forest (RF) and an optimized gradient tree boosting system (XGBoost) on distributed fog nodes. The proposed model effectiveness is assessed using an actual IoT-based dataset i.e., BoT-IoT, which includes most recent attacks found in blockchain-enabled IoT network. The results indicate, for binary attack-detection XGBoost outperforms whereas for multi-attack detection Random Forest outperforms. Overall on distributed fog nodes RF takes less time for training and testing compared to XGBoost. [ABSTRACT FROM AUTHOR]

This study aims to analyze deforestation in the Brazilian Amazon from 1999 to 2020 using machine learning techniques to assess 16 critical factors. Our approach leverages the capabilities of machine learning, particularly Random Forest, which proved to be the most accurate model in terms of determination coefficient, mean squared error, and mean absolute error. The analysis revealed that the harvested area of permanent crops is the most influential variable in predicting deforestation, followed by the area of temporary crops. Furthermore, our findings indicate a significant inverse relationship between public spending and deforestation rates. These results contribute to understanding deforestation dynamics and offer potential strategies for improving conservation efforts. • We analyze deforestation in the Brazilian Amazon from 1999 to 2020. • We employ machine learning to assess 16 critical factors. • Random Forest was the most accurate of five machine learning models. • Harvested permanent crop area dominates deforestation predicted in most models, followed by temporary crop area. • Higher public spending was linked to reduced deforestation rates, suggesting conservation interventions. [ABSTRACT FROM AUTHOR]

This study explores how well various machine learning classifiers can identify mean-motion resonances in the main belt using supervised learning. The most popular classifiers are assessed: k-Nearest Neighbours, Decision Tree, Gradient Boosting, AdaBoost, Random Forest, and Naïve Bayes. In contrast to previous studies that often relied on default ML configurations, this research conducts a detailed investigation, fine-tuning, and testing of each classifier across various parameters. The results show that simpler models, especially k-Nearest Neighbours and Decision Tree, perform better than more complex ones, particularly in terms of F 1 scores. The paper provides guides on selecting features, parameters, and training set sizes for optimal classifier performance and outlines a method for developing effective machine-learning models for asteroid classification. • Fine-tuning classifiers boosts performance in asteroid classification. • kNN and Decision Trees outperform in mean-motion resonance identification. • Naïve Bayes is effective with minimal initial dataset. [ABSTRACT FROM AUTHOR]

In the context of China's efforts to combat climate change and promote sustainable development, the solid waste treatment industry's environmental, social, and corporate governance (ESG) performance is receiving significant attention. To comprehensively assess the ESG performance of the solid waste treatment industry and identify company types, this study constructs a targeted ESG evaluation index system based on existing literature, SASB industry standards, and company reports and utilizes a random forest approach combined with K-means clustering to determine indicator weights. Based on this index system, the paper evaluates the ESG performance of 71 solid waste disposal companies (SWDCs) from 2013 to 2021 and identifies their ESG types from static and dynamic perspectives. In the static view, company types are determined based on annual ESG performance, while the dynamic view considers time-series changes to observe the evolution of company ESG types. The results show that the overall ESG performance of SWDCs falls within the 2-8-point range, indicating a noticeable high-low imbalance. Key initiatives to improve ESG performance in this industry include enhancing waste management measures, developing emergency plans, and reinforcing ESG disclosure. From a static perspective, this paper can identify companies into three categories: delayed development, single-wheel-driven, and coordinated development. Finally, from a dynamic perspective considering the time factor, companies are further subdivided into five types: continual leading, growth catch-up, slow progress, fluctuating change, and retrogressive inertia. This study not only provides targeted recommendations for different types of ESG companies but also helps various sectors of society better understand the ESG conditions of this high environmental risk industry, thereby enhancing the regulation and support for its sustainable development. [Display omitted] • ESG evaluation index system for solid waste disposal companies is proposed. • A random forest method based on k-means clustering is used to construct the index. • Comparing E, S, and G scores with industry averages to identify static ESG types. • Analyzing trends in annual static ESG types to identify dynamic ESG types. • Identify companies as three static ESG types and five dynamic ESG types. [ABSTRACT FROM AUTHOR]

Accurate identification of urban waterlogging areas and assessing waterlogging susceptibility are crucial for preventing and controlling hazards. Data-driven models are utilized to forecast waterlogging areas by establishing intricate relationships between explanatory variables and waterlogging states. This approach tackles the constraints of mechanistic models, which are frequently complex and unable to incorporate socio-economic factors. Previous research predominantly employed single-type data-driven models to predict waterlogging locations and evaluation of their effectiveness. There is a scarcity of comprehensive performance comparisons and uncertainty analyses of different types of models, as well as a lack of interpretability analysis. The chosen study area was the central area of Beijing, which is prone to waterlogging. Given the high manpower, time, and economic costs associated with collecting waterlogging information, the waterlogging point distribution map released by the Beijing Water Affairs Bureau was selected as labeled samples. Twelve factors affecting waterlogging susceptibility were chosen as explanatory variables to construct Random Forest (RF), Support Vector Machine with Radial Basis Function (SVM-RBF), Particle Swarm Optimization-Weakly Labeled Support Vector Machine (PSO-WELLSVM), and Maximum Entropy (MaxEnt). The utilization of diverse single evaluation indicators (such as F-score, Kappa, AUC, etc.) to assess the model performance may yield conflicting results. The Distance between Indices of Simulation and Observation (DISO) was chosen as a comprehensive measure to assess the model's performance in predicting waterlogging points. PSO-WELLSVM exhibited the highest performance with a DISOtest value of 0.63, outperforming MaxEnt (0.78), which excelled in identifying areas highly susceptible to waterlogging, including extremely high susceptibility zones. The SVM-RBF and RF models demonstrated suboptimal performance and exhibited overfitting. The examination of waterlogging susceptibility distribution maps predicted by the four models revealed significant spatial differences due to variations in computational principles and input parameter complexities. The integration of four WSAMs based on logistic regression has been shown to significantly decrease the uncertainty of a single data-driven model and identify the most flood-prone areas. To improve the interpretability of the data model, a geographical detector was incorporated to demonstrate the explanatory capacity of 12 variables and the process of waterlogging. Building Density (BD) exhibits the highest explanatory power in relation to explain waterlogging susceptibility (Q value = 0.202), followed by Distance to Road, Frequency of Heavy Rainstorms (FHR), DEM, etc. The interaction between BD and FHR results in a nonlinear increase in the explanatory power of waterlogging susceptibility. The presence of waterlogging susceptibility risk in the research area can be attributed to the interactions of multiple factors. • Assess urban waterlogging susceptibility using three types of data-driven WSAMs. • Evaluate comprehensive performance of WSAMs using a complex evaluation index-DISO. • WELLSVM used unlabeled clusters shows high accuracy with limited flood inventories. • Use LR-based WSAM to reduce the uncertainty of susceptibility assessment. [ABSTRACT FROM AUTHOR]

The estimation and prediction of the amount of sediment accumulated in reservoirs are imperative for sustainable reservoir sedimentation planning and management and to minimize reservoir storage capacity loss. The main objective of this study was to estimate and predict reservoir sedimentation using multilayer perceptron–artificial neural network (MLP-ANN) and random forest regressor (RFR) models in the Gibe-III reservoir, Omo-Gibe River basin. The hydrological and meteorological parameters considered for the estimation and prediction of reservoir sedimentation include annual rainfall, annual water inflow, minimum reservoir level, and reservoir storage capacity. The MLP-ANN and RFR models were employed to estimate and predict the amount of sediment accumulated in the Gibe-III reservoir using time series data from 2014 to 2022. ANN-architecture N4-100-100-1 with a coefficient of determination (R2) of 0.97 for the (80, 20) train-test approach was chosen because it showed better performance both in training and testing (validation) the model. The MLP-ANN and RFR models' performance evaluation was conducted using MAE, MSE, RMSE, and R2. The models' evaluation result revealed that the MLP-ANN model outperformed the RFR model. Regarding the train data simulation of MLP-ANN and RFR shown R2 (0.99) and RMSE (0.77); and R2 (0.97) and RMSE (1.80), respectively. On the other hand, the test data simulation of MLP-ANN and RFR demonstrated R2 (0.98) and RMSE (1.32); and R2 (0.96) and RMSE (2.64), respectively. The MLP-ANN model simulation output indicates that the amount of sediment accumulation in the Gibe-III reservoir will increase in the future, reaching 110 MT in 2030–2031, 130 MT in 2050–2051, and above 137 MTin 2071–2072. [Display omitted] • The storage capacity is the most important parameter in reservoir sediment prediction. • Forecasted data increases uncertainties in artificial neural network model prediction. • The artificial neural network model showed better performance in sediment prediction. • Uncertainty quantification determines the performance of the artificial neural network model. • Remote sensing data enhances the performance of the artificial neural network model in reservoir sedimentation prediction. [ABSTRACT FROM AUTHOR]

Co-processing recycled waste during cement production, i.e., using alternative materials such as secondary raw materials or secondary raw fuels, is widely practiced in developed countries. Alternative raw materials or fuels contain high concentrations of heavy metals and other hazardous chemicals, which might lead to the potential for dangerous heavy metals and hazardous chemicals to be transferred to clinker or cement products, resulting in exposure and emissions to people or the environment. Managing input materials and predicting which inputs affect the final concentration is essential to prevent potential hazards. We used the data of six heavy metals by input raw materials and input fuels of cement manufacturers in 2016–2017. The concentrations of Pb and Cu in cement were about 10–200 times and 4 to 200 times higher than other heavy metals (Cr, As, Cd, Hg), respectively. We profiled the influence of heavy metal concentration of each input material using the principal component analysis (PCA), which analyzed the leading causes of each heavy metal. The Random Forest (RF) ensemble model predicted cement heavy metal concentrations according to input materials. In the case of Cu, Cd, and Cr, the training performance showed R square values of 0.71, 0.71, and 0.92, respectively, as a result of predicting the cement heavy metal concentration according to the heavy metal concentration of each cement input material using the RF model, which is a machine learning model. The results of this study show that the RF model can be used to predict the amount and concentration of alternative raw materials and alternative fuels by controlling the concentration of heavy metals in cement through the concentration of heavy metals in the input materials. [Display omitted] • Heavy metal concentrations in South Korean cements were assessed. • The role of inputs materials used was an important influence factor. • RF model can expect the concentration of Cu, Cr, Cd in input materials. • This model is only applied to analyzable inputs. [ABSTRACT FROM AUTHOR]

Monitoring the health of a machining process plays a crucial role in advancing fully automated manufacturing. Faulty machining processes (abnormal process) leads to part rejection and can damage the CNC machines. In this context, data-driven machine learning (ML) models with wavelet packet transform (WPT)-based feature extraction method have gained significant importance in health monitoring. The study of different ML models, mother wavelets, and decomposition levels on real production data is important for effective process health monitoring but are not explored well in the literature. This study analyzes the performance of fifty mother wavelets, two decomposition levels (L2 and L3), and five ML models using actual production facility CNC machine dataset to achieve accurate process health monitoring. Various performance indicators obtained from the confusion matrix have been used to evaluate the ML models. The findings indicate that the Random Forest (RF) and Light-Gradient boosting machine (GBM), using the coif8 and db14 mother wavelet, at decomposition level (L3), deliver the best classification performance. Furthermore, the RF model consistently achieves good precision, sensitivity, accuracy, and F1 scores across most classes, making it a reliable and efficient option for the classification of normal and abnormal machining processes. The SVC, MLP, and CNN models, however, also exhibit competitive performance. [ABSTRACT FROM AUTHOR]

Robust quantification of vegetative biomass using satellite imagery using one or more forms of machine learning (ML) has hitherto been hindered by the extent and quality of training data. Here, we showcase how ML predictive demonstrably improves when additional training data is used. We collated field datasets of pasture biomass obtained via destructive sampling, 'C-Dax' reflective measurements and rising plate meters (RPM) from ten livestock farms across four States in Australia. Remotely sensed data from the Sentinel-2 constellation was used to retrieve aboveground biomass using a novel machine learning paradigm hereafter termed "SPECTRA-FOR" (Spectral Pasture Estimation using Combined Techniques of Random-forest Algorithm for Features Optimisation and Retrieval). Using this framework, we show that the low temporal resolution of Sentinel-2 in high latitude regions with persistent cloud cover leads to extensive gaps between cloud-free images, hindering model performance and, thus, contemporaneous ability to forecast real-time pasture biomass. By leveraging the spectral consistency between Sentinel-2 and Planet Lab SuperDove to overcome this limitation, we used ten spectral bands of Sentinel-2, four bands of Sentinel-2 as a proxy for pre-2022 SuperDove (referred to as synthetic SuperDove or SSD), and the actual SuperDove (ASD), given that SuperDove imagery has a higher resolution and more frequent passage compared with Sentinel-2. Using their respective bands as input features to SPECRA-FOR, model performance for the ten bands of Sentinel-2 were R2 = 0.87, root mean squared error (RMSE) of 439 kg DM/ha and mean absolute error (MAE) of 255 kg DM/ha, while that for SSD increased to an R2 of 0.92, RMSE of 346 kg DM/ha and MAE = 208 kg DM/ha. The study revealed the importance of robust data mining, imagery harmonisation and model validation for accurate real-time modelling of pasture biomass with ML. [Display omitted] • New method to optimise Sentinel-2 imagery using Planet Labs SuperDove data. • The near-infrared band is more sensitive for retrieving aboveground biomass. • The random forest algorithm can capture the grazing regimes of individual farms. • Model performance and evaluation rely on quality data and ample field observations. • The evaluation of the model with monthly data outperformed the varied temporal scales. [ABSTRACT FROM AUTHOR]

The ACS-NSQIP surgical risk calculator (SRC) often guides preoperative counseling, but the rarity of complications in certain populations causes class imbalance, complicating risk prediction. We aimed to compare the performance of the ACS-NSQIP SRC to other classical machine learning algorithms trained on NSQIP data, and to demonstrate challenges and strategies in predicting such rare events. Data from the NSQIP thyroidectomy module ys 2016 – 2018 were used to train logistic regression, Ridge regression and Random Forest classifiers for predicting 2 different composite outcomes of surgical risk (systemic and thyroidectomy-specific). We implemented techniques to address imbalanced class sizes and reported the area under the receiver operating characteristic (AUC) for each classifier including the ACS-NSQIP SRC, along with sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) at a 5% – 15% predicted risk threshold. Of 18,078 included patients, 405 (2.24%) patients suffered systemic complications and 1670 (9.24%) thyroidectomy-specific complications. Logistic regression performed best for predicting systemic complication risk (AUC 0.723 [0.658 – 0.778]); Random Forest with RUSBoost performed best for predicting thyroidectomy-specific complication risk (0.702; 0.674 – 0.726). The addition of optimizations for class imbalance improved performance for all classifiers. Complications are rare after thyroidectomy even when considered as composite outcomes, and class imbalance poses a challenge in surgical risk prediction. Using the SRC as a classifier where intervention occurs above a certain validated threshold, rather than citing the numeric estimates of complication risk, should be considered in low-risk patients. [ABSTRACT FROM AUTHOR]

Thyroid nodules are common; up to 67% of adults will show nodules on high-quality ultrasound, and 95% of these nodules are benign. FNA cytology is a crucial step in determining the risk of malignancy, and a false negative diagnosis at this stage delays cancer treatment. The purpose of this study is to develop a predictive model using machine learning which can identify false negative FNA results based on less-invasive clinical data. We conducted a retrospective medical record review at one academic and one community center. Inclusion criteria were thyroid nodules evaluated by ultrasound and FNA with a Bethesda II (benign) result or malignancy detected on pathology or FNA. Linear, non–linear, and ensemble models were generated with scikit-learn using 10-fold cross validation with repetition and compared with AUROC. The classification task was the prediction of malignancy using information acquired from less-invasive ultrasound and FNA. A total of 604 subjects met inclusion criteria; 38 were diagnosed with malignancy. Of all algorithms tested, a Random Forest method achieved the best AUROC (0.64) in separating benign and malignant nodules, though the improvement over other tested algorithms was not statistically significant. A Random Forest model performed better than random chance using readily available data obtained via standard evaluation of thyroid nodules. The diagnostic probability threshold of this model can be varied to minimize false positives at the cost of increasing the number of false negatives. Future studies will prospectively evaluate the model's performance. [ABSTRACT FROM AUTHOR]

• This study identify factors governing antimicrobial selectivity via interpretation of random forest models. • Random forest models produced accuracies and MCC in excess of 0.80 and 0.57, respectively. • Model interpretation was made via variable importance metrics and partial dependence plots. • Results indicated that selectivity was influenced by composition and distribution patterns of molecular charge and solubility related parameters. Host defense peptides are promising candidates for the development of novel antibiotics. To realize their therapeutic potential, high levels of target selectivity is essential. This study aims to identify factors governing selectivity via the use of the random forest algorithm for correlating peptide sequence information with their bioactivity data. Satisfactory predictive models were achieved from out-of-bag prediction that yielded accuracies and Matthew's correlation coefficients in excess of 0.80 and 0.57, respectively. Model interpretation through the use of variable importance metrics and partial dependence plots indicated that the selectivity was heavily influenced by the composition and distribution patterns of molecular charge and solubility related parameters. Furthermore, the three investigated bacterial target species (Escherichia coli , Pseudomonas aeruginosa and Staphylococcus aureus) likely had a significant influence on how selectivity was realized as there appears to be a similar underlying selectivity mechanism on the basis of charge-solubility properties (i.e. but which is tailored according to the target in question). [ABSTRACT FROM AUTHOR]

Small interfering RNAs (siRNAs) exemplify the promise of genetic medicine in the discovery of novel therapeutic modalities. Their ability to selectively suppress gene expression makes them ideal candidates for the development of oligonucleotide pharmaceuticals. Recent advancements in machine learning (ML) have facilitated the design of unmodified siRNA and efficacy prediction. However, a model trained to predict the silencing activity of siRNAs with diverse chemical modification patterns is yet to be published despite the importance of such modifications in designing siRNAs with the potential to reach the level of clinical use. This study presents the first application of ML to efficiently classify chemically modified siRNAs on the basis of sequence and chemical modification patterns alone. Three algorithms were evaluated at three classification thresholds and compared according to sensitivity, specificity, consistency of feature weights with empirical knowledge, and performance using an external validation dataset. Finally, possible directions for future research were proposed. • First ML algorithm to predict the efficacies of chemically modified siRNAs. • Three ML algorithms compared on specificity, sensitivity, and feature selection. • Models were evaluated at several efficacy thresholds to validate performance. • Random forest outperforms single-algorithm models, affirming earlier findings. • SVM, though lower in ROC AUC, provides better interpretability in feature prioritization. [ABSTRACT FROM AUTHOR]

Flow obstructed by bridge piers can increase sediment transport leading to local scour. This local scour poses a risk to the stability of bridge structures, which could lead to structural failures. There are two main approaches for evaluating the scour depth (d s) of bridge piers. The first is based on understanding hydraulic phenomena and developing relationships with properties affecting scour. The second uses data-driven soft computing models that lack physical interpretations but rely on algorithms to predict outcomes. Methods are chosen by researchers based on their goals and resources. This study aims to create innovative ensemble frameworks comprising support vector machine for regression (SVMR), random forest regression (RFR), and reduced error pruning tree (REPTree) as base learners, alongside bagging regression tree (BRT) and stochastic gradient boosting (SGB) as meta learners. These ensembles were developed to analyse maximum scour depths (d sm) in clear water conditions, utilizing 35 literature's experimental data published in last 63 years. The performance of each machine learning (ML) approach was assessed using statistical performance indicators. The proposed model was also compared with top six empirical equations with strong predictive ability. Results show that among these empirical equations, the equation from Nandi and Das (2023) performs best. Performance evaluation considering training, testing, and the entire dataset, SGB (REPTree), BRT(SVMR-PUK), and SGB (REPTree) exhibited the highest performance, securing the top rank among all ML models and empirical equations. Sensitivity analysis identified sediment gradation and flow intensity as the most influential variables for predicting d sm during both training and testing phases, respectively. • Predict maximum clearwater scour (d sm), by metadata analysis of 850 datasets from 35 literatures. • Compare six most promising empirical equations alongwith novel ML techniques. • Among all models, SGB(REPTree), BRT(SVMR-PUK), and RFR perform best for training, testing and all datasets, respectively. • Sensitivity analysis confirms sediment gradation and flow intensity as the most influential parameters. • nalysis identifies SGB(RFR) model with lowest and Jain and Fischer (1979) equation with highest prediction uncertainty. [ABSTRACT FROM AUTHOR]

Data-driven machine learning approaches are promising to substitute physically based groundwater numerical models and capture input-output relationships for reducing computational burden. But the performance and reliability are strongly influenced by different sources of uncertainty. Conventional researches generally rely on a stand-alone machine learning surrogate approach and fail to account for errors in model outputs resulting from structural deficiencies. To overcome this issue, this study proposes a flexible integrated Bayesian machine learning modeling (IBMLM) method to explicitly quantify uncertainties originating from structures and parameters of machine learning surrogate models. An Expectation-Maximization (EM) algorithm is combined with Bayesian model averaging (BMA) to find out maximum likelihood and construct posterior predictive distribution. Three machine learning approaches representing different model complexity are incorporated in the framework, including artificial neural network (ANN), support vector machine (SVM) and random forest (RF). The proposed IBMLM method is demonstrated in a field-scale real-world "1500-foot" sand aquifer, Baton Rouge, USA, where overexploitation caused serious saltwater intrusion (SWI) issues. This study adds to the understanding of how chloride concentration transport responds to multi-dimensional extraction-injection remediation strategies in a sophisticated saltwater intrusion model. Results show that most IBMLM exhibit r values above 0.98 and NSE values above 0.93, both slightly higher than individual machine learning, confirming that the IBMLM is well established to provide better model predictions than individual machine learning models, while maintaining the advantage of high computing efficiency. The IBMLM is found useful to predict saltwater intrusion without running the physically based numerical simulation model. We conclude that an explicit consideration of machine learning model structure uncertainty along with parameters improves accuracy and reliability of predictions, and also corrects uncertainty bounds. The applicability of the IBMLM framework can be extended in regions where a physical hydrogeologic model is difficult to build due to lack of subsurface information. • An integrated Bayesian machine learning modeling method is newly developed. • Uncertainties from surrogate model structures and parameters are explicitly quantified. • Most IBMLM exhibit r values above 0.98 and NSE values above 0.93. • Saltwater intrusion response to aquifer remediation strategies is quickly predicted. • Extraction-injection along with a cutoff barrier works effectively to control SWI. [ABSTRACT FROM AUTHOR]

The abundance and diversity of the microflora in a complex environment such as soil is everchanging. Mica mining has led to metalloid poisoning and changes in soil biogeochemistry affecting the overall produce and leading to toxic dietary exposure. The study focuses on two prominent stressors acidity and arsenic, in mining-contaminated agricultural locations. Soil samples were collected from agricultural fields at a distance of 50 m (zone 1) and 500 m (zone 2) from active mines. Mean arsenic concentration was higher in zone 1 and pH was lower. Geostatistical and self-organizing maps were employed to report that the pattern of localization of soil acidity and arsenic content is similar indicating a causal relationship. Cluster and principal component analysis were further used to materialize a negative effect of soil acidity fractions and arsenic labile pool on soil enzymatic activity (fluorescein diacetate, dehydrogenase, β-1,4-glucosidase, phosphatase, and urease), respiration and Microbial biomass carbon. Soil metagenomic analysis revealed significant differences in the abundance of microbial populations with zone 1 (contaminated zone) having lower alpha and beta diversity. Finally, the efficacy of several machine-learning tools was tested using Taylor diagrams and an effort was made to select a potent algorithm to predict the causal stressors responsible for depreciating soil microbial health. Random Forrest had superior predictive power based on numerical evidence and was therefore chosen as the best-fitted model. The aforementioned insights into soil microbial health and sustenance in stressed conditions can be beneficial for predicting remedial strategies and practicing sustainable agriculture. • Mica mining triggers acidity & arsenic stress, harming agricultural yield and human health in mine adjacent localities. • Acidity of soil mediates solubilization and mobility of arsenic. • Acidity and microbial stress led to reduced microbial metabolism and shifts in microbial diversity structure. • Random forest algorithm is efficient in predicting soil quality and health from physicochemical stressor inputs. [ABSTRACT FROM AUTHOR]

Non-targeted analysis (NTA) has great potential to screen emerging contaminants in the environment, and some studies have conducted in-depth investigation on environmental samples. Here, we used a NTA workflow to identify emerging contaminants in used tire particle (TP) leachates, followed by quantitative prediction and toxicity assessment based on hazard scores. Tire particles were obtained from four different types of automobiles, representing the most common tires during daily transportation. With the instrumental analysis of TP leachates, a total of 244 positive and 104 negative molecular features were extracted from the mass data. After filtering by a specialized emerging contaminants list and matching by spectral databases, a total of 51 molecular features were tentatively identified as contaminants, including benzothiazole, hexaethylene glycol, 2-hydroxybenzaldehyde, etc. Given that these contaminants have different mass spectral responses in the mass spectrometry, models for predicting the response of contaminants were constructed based on machine learning algorithms, in this case random forest and artificial neural networks. After five-fold cross-validation, the random forest algorithm model had better prediction performance (MAE CV = 0.12, Q2 = 0.90), and thus it was chosen to predict the contaminant concentrations. The prediction results showed that the contaminant at the highest concentration was benzothiazole, with 4,875 μg/L in the winter tire sample. In addition, the joint toxicity assessment of four types of tires was conducted in this study. According to different hazard levels, hazard scores increasing by a factor 10 were developed, and hazard scores of all the contaminants identified in each TP leachate were summed to obtain the total hazard score. All four tires were calculated to have relatively high risks, with winter tires having the highest total hazard score of 40,751. This study extended the application of NTA research and led to the direction of subsequent targeting studies on highly concentrated and toxic contaminants. • Non-targeted analysis with extended applications were conducted for tire leachates. • Comprehensive information on emerging contaminants in the leachate of used tires was provided. • Random Forest model demonstrated better potential for predicting contaminant concentrations. • Used winter tires might release more hazardous contaminants. [ABSTRACT FROM AUTHOR]

Soil loss is an environmental concern of global importance. Accurate simulation of soil loss in small watersheds is crucial for protecting the environment and implementing soil and water conservation measures. However, predicting soil loss while meeting the criteria of high precision, efficiency, and generalizability remains a challenge. Therefore, this study first used three machine learning (ML) algorithms, namely, random forest (RF), support vector machine (SVM), and artificial neural network (ANN) to develop soil loss models and predict soil loss rates (SLRs). These soil loss models were constructed using field observation data with an average SLR of 1756.48 t/km2 from rainfall events and small watersheds in the hilly-gully region of the Loess Plateau, China. During training, testing and generalizability stages, the average coefficients of determination from the RF, SVM, and ANN models were 0.903, 0.860, and 0.836, respectively. Similarly, the average Nash-Sutcliffe coefficients of efficiency from the RF, SVM and ANN models were 0.893, 0.791 and 0.814, respectively. These results indicated that MLs have superior predictive performance and generalizability, and broad prospects for predicting SLRs. This study also demonstrated that the RF model outperformed better than the SVM and ANN models. Therefore, the RF model was used to simulate the SLR of each small watershed in the Chabagou watershed. Our results showed the four-year (2017–2020) average annual SLR of the small watersheds ranged from 0.73 to 1.63 × 104 t/(km2∙a) in the Chabagou watershed. Additionally, the results also indicated the SLR of small watersheds under the rainstorm event with a 100-year recurrence interval was 4.4–51.3 times that of other rainfall events.Furthermore, this study confirmed that bare land was the predominant source of soil loss in the Chabagou watershed, followed by cropland land and grassland. This study helps to provide the theoretical basis for deploying soil and water conservation measures to realize the sustainable utilization of soil resources in the future. [Display omitted] • Three machine learning algorithms were first used to build soil loss models. • Random forest demonstrated the most favorable performance for soil loss modeling. • Soil loss caused by rainstorms was 4.4–51.3 times that of other rainfall events. • Bare land was the predominant source of soil loss. [ABSTRACT FROM AUTHOR]

Biodiversity datasets with high spatial resolution are critical prerequisites for river protection and management decision-making. However, traditional morphological biomonitoring is inefficient and only provides several site estimates, and there is an urgent need for new approaches to predict biodiversity on fine spatial scales throughout the entire river systems. Here, we combined the environmental DNA (eDNA) and remote sensing (RS) technologies to develop a novel approach for predicting the spatial distribution of aquatic insects with high spatial resolution in a disturbed subtropical Dongjiang River system of southeast China. First, we screened thirteen RS-based vegetation indices that significantly correlated with the eDNA-inferred richness of aquatic insects. In particular, the green normalized difference vegetation index (GNDVI) and normalized difference red-edge2 (NDRE2) were closely related to eDNA-inferred richness. Second, using the gradient boosting decision tree, our data showed that the spatial pattern of eDNA-inferred richness could achieve a high spatial resolution to 500 m reach and accurate prediction of more than 80%, and the prediction efficiency of the headwater streams (Strahler stream order = 1) was slightly higher than the downstream (Strahler stream order >1). Third, using the random forest algorithm, the spatial distribution of aquatic insects could reach a prediction rate of over 70% for the presence or absence of specific genera. Overall, this study provides a new approach to achieving high spatial resolution prediction of the distribution of aquatic insects, which supports decision-making on river diversity protection under climate changes and human impacts. [Display omitted] • EDNA and remote sensing datasets are combined to predict species distribution. • Thirteen RS-based vegetation indices are closely related to eDNA-inferred richness. • Aquatic insects can achieve spatial distribution prediction within 500 m reach. [ABSTRACT FROM AUTHOR]

Ecological restoration projects in the Tibetan Plateau aimed to reverse ecosystem degradation and safeguard the fragile alpine ecological environment. However, it is still being determined if the vegetation restoration is successful on a large scale or reaches the expected magnitude, restricting our ability to adapt practices to maximise the benefit. With multiple vegetation indices (VIs: NDVI, LAI, and GPP) from satellite observations and random forest machine-learning models, we performed an attribution study on vegetation growth trends caused by climate change and human activities. Then, we further explored the relationship between vegetation growth and ecological projects and human footprint without the influence of climate. The results showed that climatic change was a relatively strong driver of vegetation growth. The positive contributions of ecological restoration occurred only in half of the plateau due to the increased human footprint. Vegetation enhancement resulting from ecological restoration occurred in 13.1%–23.1% of the plateau. Among these values, ecological restoration counteracted the negative climate effects in 4.7%–8.3% of the plateau (about half of the negative climate effect area). In forest and grassland protection areas, the ecological restoration was more successful. The study provides a better understanding of the role of ecological projects in vegetation restoration and potential threats to its effectiveness. This is essential to improve future restoration projects. • Climate change primarily driven vegetation growth trends in the Tibetan Plateau. • The positive contributions of ecological restoration occurred only in half of the plateau due to continued human pressures. • Vegetation enhancement resulting from ecological restoration occurred in 13.1%–23.1% of the plateau. • Ecological restoration prevented nearly half of the areas subjected to adverse climate forces from experiencing degradation. • In forest and grassland protection areas, the ecological restoration was more successful. [ABSTRACT FROM AUTHOR]

Tissue-to-blood partition coefficients (P tb) are crucial for assessing the distribution of chemicals in organisms. Given the lack of experimental data and laborious nature of experimental methods, there is an urgent need to develop efficient predictive models. With the help of machine learning algorithms, i,e., random forest (RF), and artificial neural network (ANN), this study developed multi-task (MT) models that can simultaneously predict P tb values for various mammalian tissues, including liver, muscle, brain, lung, and adipose. Single-task (ST) models using partial least squares regression, RF, and ANN algorithms for each endpoint were established for comparison. Overall, the performances of MT models were superior to those of ST models. The MT model using ANN algorithms showed the highest prediction accuracy with determination coefficients ranging from 0.704 to 0.886, root mean square errors between 0.223 and 0.410, and mean absolute errors ranging from 0.178 to 0.285 log units. Results showed that lipophilicity and polarizability of molecules significantly influence their partition behavior in organisms. Applicability domains (ADs) of the models were characterized by weighted molecular similarity density, and weighted inconsistency in molecular activities of structure-activity landscapes. When constrained by ADs, the models displayed enhanced predictive accuracy, making them valuable tools for the risk assessment and management of chemicals. [Display omitted] • Multi-task machine learning models were developed to predict tissue-to-blood partition coefficients for mammalian tissues. • Multi-task models can improve model prediction accuracy by capturing shared information across similar tasks. • Lipophilicity and polarizability of molecules significantly influence their partition behavior in organisms. [ABSTRACT FROM AUTHOR]

Computational models of fluid flows based on the Reynolds-Averaged Navier–Stokes (RANS) equations supplemented with a turbulence model are the gold standard in engineering applications. A plethora of turbulence models and related variants exist, none of which are fully reliable outside the range of flow configurations for which they were calibrated. Thus, the choice of a suitable turbulence closure largely relies on subjective expert judgement and engineering know-how. In this article, we propose a data-driven methodology for combining the solutions of a set of competing turbulence models. The individual model predictions are linearly combined for providing an ensemble solution accompanied by estimates of predictive uncertainty due to the turbulence model choice. First, for a set of training flow configurations we assign high weights to component models in the regions where they perform best, and vice versa, based on a measure of distance between high-fidelity data and individual model predictions. The model weights are then mapped into a space of features, representative of the local flow physics, and regressed using a Random Forests (RF) algorithm. Finally, the RF regressor is used to infer spatial distributions of the model weights for unseen configurations. Predictions of new cases are constructed as a convex linear combination of the underlying models solutions, while the between-model variance provides information about regions of high model uncertainty. The method is demonstrated for a class of flows through the compressor cascade NACA65 V103 at R e ≃ 3 × 10 5. The results show that the aggregated solution outperforms the accuracy of individual models for the quantity used to inform the RF regressor, and performs well for other quantities well-correlated with the previous one. The estimated uncertainty intervals are generally consistent with the target high-fidelity data. The present approach then represents a viable methodology for a more objective selection and combination of alternative turbulence models in configurations of interest for engineering practice. [ABSTRACT FROM AUTHOR]

The regulation of olive cultivar and geographical origin is a requirement for the global extra virgin olive oil market, due to its significant impact on consumer choice. Our work involves obtaining a promising marker parameter for cultivar and geographical origin that can be used to verify declared labels. The effects of these factors on the physicochemical parameters and composition of monovarietal extra virgin olive oil (MEVOO) from Algeria were studied. Thirteen olive fruit varieties were analyzed using different physicochemical methods, including phenolic and fatty acid composition. Five classification techniques, random forests (RForest), gradient boosted trees (GBoost), Naïve Bayes (NBayes), logistic regression (LRegression) and decision tree (DTree), were applied and their results were compared. The best validation accuracy of 91.7 % was achieved with DT classification through a feature selection procedure using a genetic algorithm (GA). These results demonstrate the effective use of machine learning techniques to rapidly classify different Algerian varieties based on their compositional fingerprints. [Display omitted] • 45 monovarietal extra virgin olive oils of 13 cultivars from 17 regions were studied. • Machine learning and feature selection were used to discriminate and classify olive oils. • K 232 and linolenic acid are indicative parameters for classifying olive oil cultivars. [ABSTRACT FROM AUTHOR]

Bisphenol A (BPA) is easily enriched in many human-disturbed watersheds, particularly lakes with poor water mobility, which is posing a threat to aquatic biota. While previous studies have focused on the concentration of BPA in water and its toxicity to aquatic organisms, a small amount of measured data is not enough to reveal the temporal and spatial distribution and threats of BPA, and estimate the ecological risk in watersheds. Therefore, we collected 164 measured BPA data points from Taihu Lake to develop machine learning models using random forest (RF), support vector machine (SVM) and least square regression (LSR) and created month-by-month watershed prediction maps in temperate lakes to estimate the spatiotemporal distribution and threats of BPA. Due to RF's superior robustness to noisy data, the RF model exhibits the best performance among the three algorithms. The RF model showed acceptable predictive performance on the modeling dataset (coefficients of determination and root-mean-square error for the training set were 0.927 and 17.499, respectively, and 0.607, 39.645 for the validation set, respectively). The maps indicated that areas susceptible to anthropogenic activities were more severely polluted by BPA, and rainy climate may favor the migration of BPA to aquatic ecosystems. The model was also applied to predict 42 data points of BPA collected from Dianchi Lake, and the results showed that most predicted data were within a factor of 10 of the measured data, but the prediction accuracy of the model has declined. The ecological risks in the two lakes were evaluated and attention should be paid to the regions with higher risks. Our study provided a novel idea for comprehensive monitoring of an unconventional trace pollutant with endocrine disrupting effects in aquatic ecosystems and analyzing their spatiotemporal distribution, which will contribute to the scientific assessment of the ecological risk of BPA. [Display omitted] • Random forest model predicted the concentration of BPA. • Watershed prediction maps showed the spatiotemporal distribution of BPA in temperate lakes. • Factors that affected the spatiotemporal distribution of BPA were analyzed. • Regions with higher ecological risks were identified. [ABSTRACT FROM AUTHOR]

The lack of readily available methods for estimating high-resolution near-surface relative humidity (RH) and the incapability of weather stations to fully capture the spatiotemporal variability can lead to exposure misclassification in studies of environmental epidemiology. We therefore aimed to predict German-wide 1 × 1 km daily mean RH during 2000–2021. RH observations, longitude and latitude, modelled air temperature, precipitation and wind speed as well as remote sensing information on topographic elevation, vegetation, and the true color band composite were incorporated in a Random Forest (RF) model, in addition to date for capturing the temporal variations of the response-explanatory variables relationship. The model achieved high accuracy (R2 = 0.83) and low errors (Root Mean Square Error (RMSE) of 5.07%, Mean Absolute Percentage Error (MAPE) of 5.19% and Mean Percentage Error (MPE) of - 0.53%), calculated via ten-fold cross-validation. A comparison of our RH predictions with measurements from a dense monitoring network in the city of Augsburg, South Germany confirmed the good performance (R2 ≥ 0.86, RMSE ≤ 5.45%, MAPE ≤ 5.59%, MPE ≤ 3.11%). The model displayed high German-wide RH (22y-average of 79.00%) and high spatial variability across the country, exceeding 12% on yearly averages. Our findings indicate that the proposed RF model is suitable for estimating RH for a whole country in high-resolution and provide a reliable RH dataset for epidemiological analyses and other environmental research purposes. • We estimated daily 1 × 1 km near-surface relative humidity (RH) in Germany, 2000–2021. • The random forest model achieved good performance (R2 = 0.83, RMSE = 5.07%). • Validation with external data confirmed the model's high accuracy and low errors. • We propose an RH modeling process generalizable to other study domains/countries. • We provide reliable and highly-resolved RH data for epidemiological studies. [ABSTRACT FROM AUTHOR]

Riparian vegetation composition and channel morphology are susceptible to long-term alterations caused by external stressors, including climate-change-induced droughts and engineered infrastructures. The objectives of this study were to (1) quantify trends in riparian vegetation and channel/floodplain morphology over large spatial (∼290 km) and temporal scales (∼30 years) and (2) investigate the relationships between hydroclimatic drivers and changes in riparian vegetation and channel morphology. We implemented a random forest classifier via a machine learning technique in Google Earth Engine. The study area was a 290 km reach of the Rio Grande located in New Mexico, USA. We used the combination of remotely sensed data and products (e.g., Landsat imagery, Normalized Difference Vegetation Index (NDVI), and land cover) to characterize vegetation, vegetation cover changes, and river morphology shifts from 1984 to 2020. The trend analysis revealed increased vegetated areas and NDVI (0.0004/yr) during long-term drought. The channel experienced a reduction in width associated with vegetation encroachment and the formation of stable vegetated islands. The streamflow hydrograph characteristics were positively correlated with vegetation cover and channel morphology. Our study contributes novel insights into the long-term riparian ecosystem dynamics under drought stress, informing drought impact mitigation and ecosystem management in arid and semi-arid regions. • Changes in riparian vegetation, channel morphology, and hydrology have cascading impacts on the river's ecological function. • The vegetated area has increased, and the channel has narrowed along the Rio Grande during persistent drought. • Maintenance of channel width is challenging once vegetation is established. [ABSTRACT FROM AUTHOR]

Farms use large quantities of fertilizers from many sources, making quality control a challenging task, as the traditional wet-chemistry analyses are expensive, time consuming and not environmentally-friendly. As an alternative, this work proposes the use of portable X-ray fluorescence (pXRF) spectrometry and machine learning algorithms for rapid and low-cost estimation of macro and micronutrient contents in mineral and organic fertilizers. Four machine learning algorithms were tested. Whole (i.e., as delivered by the manufacturer) (CP) and ground (AQ) samples (429 in total) were analyzed to test the effect of fertilizer granulometry in prediction performance. Model validation indicated highly accurate predictions of macro (N: R2 = 0.92; P: 0.97; K: 0.99; Ca: 0.94, Mg: 0.98; S: 0.96) and micronutrients (B: 0.99; Cu: 0.99; Fe: 0.98; Mn: 0.91; Zn: 0.94) for both organic and mineral fertilizers. RPD values ranged from 2.31 to 9.23 for AQ samples, and Random Forest and Cubist Regression were the algorithms with the best performances. Even samples analyzed as they were received from the manufacturer (i.e., no grinding) provided accurate predictions, which accelerate the confirmation of nutrient contents contained in fertilizers. Results demonstrated the potential of pXRF data coupled with machine learning algorithms to assess nutrient composition in both mineral and organic fertilizers with high accuracy, allowing for clean, fast and accurate quality control. Sensor-driven quality assessment of fertilizers improves soil and plant health, crop management efficiency and food security with a reduced environmental footprint. [Display omitted] • Quality control of organic and mineral fertilizers can be accurately done via pXRF. • Sample preprocessing did not affect results significantly. • Random Forest and Cubist Regression achieved the best prediction performances. • Nutrient content assessment in fertilizers is possible via a green method. [ABSTRACT FROM AUTHOR]

The sight-threatening sulfur mustard (SM) induced ocular injury presents specific symptoms in each clinical stage. The acute injury develops in all exposed eyes and may heal or deteriorate into chronic late pathology. Early detection of eyes at risk of developing late pathology may assist in providing unique monitoring and specific treatments only to relevant cases. In this study, we evaluated a machine-learning (ML) model for predicting the development of SM-induced late pathology based on clinical data of the acute phase in the rabbit model. Clinical data from 166 rabbit eyes exposed to SM vapor was used retrospectively. The data included a comprehensive clinical evaluation of the cornea, eyelids and conjunctiva using a semi-quantitative clinical score. A random forest classifier ML model, was trained to predict the development of corneal neovascularization four weeks post-ocular exposure to SM vapor using clinical scores recorded three weeks earlier. The overall accuracy in predicting the clinical outcome of SM-induced ocular injury was 73%. The accuracy in identifying eyes at risk of developing corneal neovascularization and future healed eyes was 75% and 59%, respectively. The most important parameters for accurate prediction were conjunctival secretion and corneal opacity at 1w and corneal erosions at 72 h post-exposure. Predicting the clinical outcome of SM-induced ocular injury based on the acute injury parameters using ML is demonstrated for the first time. Although the prediction accuracy was limited, probably due to the small dataset, it pointed out towards various parameters during the acute injury that are important for predicting SM-induced late pathology and revealing possible pathological mechanisms. • A machine learning model predicts the clinical outcome of sulfur mustard-induced ocular injury. • Clinical scores of the acute injury in the rabbit model were used to train a random forest classifier. • The overall accuracy in predicting the clinical outcome is 73%. • The algorithm pointed out the important clinical parameters for accurate prediction. [ABSTRACT FROM AUTHOR]

Flash floods are one of the worst natural disasters, causing massive economic losses and many deaths. Creating a flood susceptibility map (FSM) that pinpoints the areas most at risk of flooding is a crucial non-structural solution for managing floods. This study aimed to assess the efficacy of combinations of the random forest (RF) model with three biology-inspired metaheuristic algorithms, namely invasive weed optimization (IWO), slime mould algorithm (SMA), and satin bowerbird optimization (SBO), for flood susceptibility mapping in Estahban town, Iran. Initially, synthetic-aperture radar (SAR) (Sentinel-1) and optical (Landsat-8) satellite images were integrated to monitor the flooded areas during the July 2022 monsoon in the study area. A dataset of 509 flood occurrence points was created to identify flood-prone areas using remote sensing techniques, considering the monitored flood areas. The dataset also included twelve flood-related criteria: topography, land cover, and climate. The holdout method was employed for modeling, with a ratio of 70:30 used for the train/test split. Data pre-processing techniques were conducted to improve model performance, including determining criteria importance and addressing multicollinearity issues using certainty factor (CF), multicollinearity, and information gain ratio (IGR) methods. Then FSM was prepared using RF, RF-IWO, RF-SBO, and RF-SMA models. The findings of this research revealed that the RF-IWO model was the best predictive model of flood susceptibility modeling, with root-mean-square-error (RMSE) (0.211 and 0.0.27), mean-absolute-error (MAE) (0.103 and 0.15), and coefficient-of-determination (R2) (0.821 and 0.707) in the training and testing phases, respectively. Receiver operating characteristic (ROC) curve analysis of FSM revealed that the most accurate models were the RF-IWO (area under the curve (AUC) = 0.983), RF-SBO (AUC = 0.979), RF-SMA (AUC = 0.963), and RF (AUC = 0.959), respectively. Integrating biology-inspired computing algorithms with machine learning algorithms presents a novel approach to enhancing the accuracy of FSMs. [Display omitted] • Detected monsoon floods and identified vulnerable areas. • Integrated optical and radar satellite images to improve flood monitoring accuracy. • The biology-inspired metaheuristic algorithms increased flood prediction accuracy. • Strong spatial association of flood probability (0.959–0.983) among models. [ABSTRACT FROM AUTHOR]

According to frontier settlement theory, physical topography is a critical factor that shapes a region's psychological landscape. To determine whether physical topography also impacts generalized trust, we conducted an exploratory study to investigate the relationship between mountainousness and individual's generalized trust (N = 26,910) in China. The multilevel modeling analysis revealed positive associations between the standard deviation of elevation and generalized trust. Random Forest analyses confirmed mountainousness indicators as significant predictors of generalized trust and also uncovered a complex nonlinear relationship between these indicators and generalized trust. Our results revealed that individuals residing in China's mountainous areas are more prone to trust others. • The present research investigates the association between mountainousness and generalized trust in China. • The standard deviation of elevation is positively correlated with generalized trust. • Random forest analyzes reveal nonlinear relationships between mountainousness indicators and generalized trust. • Individuals residing in China's mountainous areas are more prone to trust others. [ABSTRACT FROM AUTHOR]