Search

The Balangu (Lallemantia iberica) species have a high gastronomical impact in the Middle East and Balkan region. It is widely used in the local food industry, such as confectionery, edible oil, and protein food. In this study, 49 ecotypes were collected from different regions of Iran. 37 agronomic traits were measured during the growing season and at harvest time. To find the correlation between the grain yield per unit area, grain yield per single plant (GYSP), oil percent (OP), and protein percent (PP) with other measured traits, which these were utilized as the labels of different machine learning (ML) procedures including Linear Regression (LR), Support Vector Regression (SVR), Random Forest Regression (RFR), and Gradient Boosting Decision Tree Regression (GBDTR). It was observed that there is a linear relationship between the measured agronomic traits and the considered labels. So, the LR, RFR, and GBDTR models showed the lowest mean absolute error, mean square error, and root mean square error than SVR models and good prediction ability of the test data. Although, the RFR and GBDTR have naturally lower bias than other methods in this study, but the GBDTR scheme is preferred because of the over-fitting shortcoming of the RFR technique. The GBDTR method showed better results rather than the other ML regression methods according to the RMSE 3.302, 0.040, 0.028, and 0.060 for GYUA, GYSP, OP, and PP, respectively. [ABSTRACT FROM AUTHOR]

Hemorrhagic events are the main focus of attention during antithrombosis therapy in patients with coronary artery disease (CAD). This study aims to investigate the potential of using photoplethysmography (PPG) and machine learning techniques to assess hemorrhagic risk in patients with CAD. A total of 1638 patients with CAD were enrolled from January 2018 to October 2019, among which 114 patients were observed to have at least one positive event. Importantly, 102 patients with 9933 records were finally retained for analysis in this study. Participants were required to collect data using the portable PPG acquisition device and the specially designed Android APP. The data was collected and uploaded to a remote server. Based on collected PPG signals, we extracted features in a total of 30 dimensions from time-domain, frequency-domain, and wavelet packet decomposition. Logistic regression, support vector regression, random forest, and XGBoost regression models were established to achieve hemorrhagic risk evaluation, and then, their performances were compared. In total, 10 features extracted from PPG showed statistical significance (p < 0.01) between negative and positive groups. The newly established XGBoost model performed best in the hemorrhagic risk evaluation experiment, wherein the mean area under the curve (AUC) with tenfold cross-validation was 0.762 ± 0.024 and the sensitivity and specificity were 0.679 ± 0.051 and 0.714 ± 0.014, respectively. We established a data acquisition system for PPG signal collection, and demonstrated that a set of features extracted from PPG and the proposed machine learning model are promising in the evaluation of hemorrhagic risk among patients with CAD. In comparison with the traditional HAS-BLED score, the proposed method can obtain the quantitative risk prediction probability from a single PPG record, which has the advantages of dynamics and continuity, and can provide timely feedback for doctors' antithrombotic treatment, which is of great significance for doctors to quickly determine the effectiveness of the treatment and adjust the timely treatment plans accordingly. [ABSTRACT FROM AUTHOR]

Preterm birth is a global public health problem with a significant burden on the individuals affected. The study aimed to extend current research on preterm birth prognostic model development by developing and internally validating models using machine learning classification algorithms and population-based routinely collected data in Western Australia. The longitudinal retrospective cohort study involved all births in Western Australia between 1980 and 2015, and the analytic sample contains 81,974 (8.6%) preterm births (< 37 weeks of gestation). Prediction models for preterm birth were developed using regularised logistic regression, decision trees, Random Forests, extreme gradient boosting, and multi-layer perceptron (MLP). Predictors included maternal socio-demographics and medical conditions, current and past pregnancy complications, and family history. Class weight was applied to handle imbalanced outcomes and stratified tenfold cross-validation was used to reduce overfitting. Close to half of the preterm births (49.1% at 5% FPR, 95% CI 48.9%,49.5%) were correctly classified by the best performing classifier (MLP) for all women when current pregnancy information was available. The sensitivity was boosted to 52.7% (95% CI 52.1%,53.3%) after including past obstetric history in a sub-population of births from multiparous women. Around half of the preterm birth can be identified antenatally at high specificity using population-based routinely collected maternal and pregnancy data. The performance of the prediction models depends on the available predictor pool that is individual and time specific. [ABSTRACT FROM AUTHOR]

Karst rocky desertification (KRD) has become one of the most serious ecological and environmental problems in karst areas. At present, mapping KRD with a high accuracy and on a large scale is still a difficult problem in the control of KRD. In this study, a random forest (RF) based on maximum information coefficient and correlation coefficient feature selection is proposed to predict KRD. Nine predictors stood out as feature factors to estimate KRD. Rock exposure was the most important predictor, followed by fractional vegetation cover for the prediction of KRD processes. The kappa and classification accuracy indexes were to evaluate the performance of the model. We recorded overall accuracy rate and kappa index values of 94.7% and 0.92 for the testing datasets respectively. The RF model was then used to predict the KRD in 2001, 2011, 2016, and 2020, and it was found that the KRD in the study area has exhibited a positive trend of improvement. Therefore, the use of multisource remote sensing data combined with the RF model can obtain better prediction results of KRD, thereby providing a new idea for large-scale estimation of the KRD in peak-cluster depression. [ABSTRACT FROM AUTHOR]

Validating scales for clinical use is a common procedure in medicine and psychology. Through the application of computational methods, we present a new strategy for estimating construct validity and criterion validity. XGBoost, Random Forest and Support-Vector machine learning algorithms were employed in order to make predictions based on the pattern of participants' responses by systematically controlling computational experiments with artificial experiments whose results are guaranteed. According to these findings, these approaches are capable of achieving construct and criterion validity and therefore could provide an additional layer of evidence to traditional validation approaches. In particular, this study examined the extent to which measured items are inferable by theoretically related items, as well as the extent to which the information carried by a given construct can be translated into other theoretically compatible normative scales based on other constructs (thereby providing information about construct validity); as well as the replicability of clinical decision rules on several partitions (thereby providing information about criterion validity). [ABSTRACT FROM AUTHOR]

As the amount of data on farms grows, it is important to evaluate the potential of artificial intelligence for making farming predictions. Considering all this, this study was undertaken to evaluate various machine learning (ML) algorithms using 52-year data for sheep. Data preparation was done before analysis. Breeding values were estimated using Best Linear Unbiased Prediction. 12 ML algorithms were evaluated for their ability to predict the breeding values. The variance inflation factor for all features selected through principal component analysis (PCA) was 1. The correlation coefficients between true and predicted values for artificial neural networks, Bayesian ridge regression, classification and regression trees, gradient boosting algorithm, K nearest neighbours, multivariate adaptive regression splines (MARS) algorithm, polynomial regression, principal component regression (PCR), random forests, support vector machines, XGBoost algorithm were 0.852, 0.742, 0.869, 0.915, 0.781, 0.746, 0.742, 0.746, 0.917, 0.777, 0.915 respectively for breeding value prediction. Random forests had the highest correlation coefficients. Among the prediction equations generated using OLS, the highest coefficient of determination was 0.569. A total of 12 machine learning models were developed from the prediction of breeding values in sheep in the present study. It may be said that machine learning techniques can perform predictions with reasonable accuracies and can thus be viable alternatives to conventional strategies for breeding value prediction. [ABSTRACT FROM AUTHOR]

Different modeling techniques must be applied to manage production and statistical estimation to predict the expected harvest. By calculating advanced production methods and the rational valuation of different factors, we can accurately capture the variety of growth characteristics and the expected yield. This paper obtained 32 feature variables related to melons, including phenological features, shape features, and color features. The Gradient Boosted Decision Tree (GBDT) network and the Grid Search (GS) hyperparameter seeking method was applied to calculate the degree of importance of all melon fruits' characteristics and construct prediction models for three expected harvest indexes of melon yield, sugar content, and endocarp hardness. To facilitate growers to carry out prediction and estimation in the field without destroying the melon fruits. The reduced feature variables were selected as inputs. The GBDT model was used to provide a significant advantage in prediction compared to both Random Forest (RF) and Support Vector Regression (SVR) methods. In addition, to verify the feasibility of using only reduced feature variables as input for the evaluation work, this study also compares the predictive effects of the model when all feature variables and only reduced feature variables are used. The GBDT prediction model proposed in this paper predicted melon yield, sugar content, and hardness using reduced features as input, and the model R2 could reach more than 90%. Therefore, this method can effectively help growers carry out early non-destructive inspection and growth prediction of melons in the field. [ABSTRACT FROM AUTHOR]

Accurate identification of drug-targets in human body has great significance for designing novel drugs. Compared with traditional experimental methods, prediction of drug-targets via machine learning algorithms has enhanced the attention of many researchers due to fast and accurate prediction. In this study, we propose a machine learning-based method, namely XGB-DrugPred for accurate prediction of druggable proteins. The features from primary protein sequences are extracted by group dipeptide composition, reduced amino acid alphabet, and novel encoder pseudo amino acid composition segmentation. To select the best feature set, eXtreme Gradient Boosting-recursive feature elimination is implemented. The best feature set is provided to eXtreme Gradient Boosting (XGB), Random Forest, and Extremely Randomized Tree classifiers for model training and prediction. The performance of these classifiers is evaluated by tenfold cross-validation. The empirical results show that XGB-based predictor achieves the best results compared with other classifiers and existing methods in the literature. [ABSTRACT FROM AUTHOR]

This study aims to provide a transferable methodology in the context of sport performance modelling, with a special focus to the generalisation of models. Data were collected from seven elite Short track speed skaters over a three months training period. In order to account for training load accumulation over sessions, cumulative responses to training were modelled by impulse, serial and bi-exponential responses functions. The variable dose-response (DR) model was compared to elastic net (ENET), principal component regression (PCR) and random forest (RF) models, while using cross-validation within a time-series framework. ENET, PCR and RF models were fitted either individually ( M I ) or on the whole group of athletes ( M G ). Root mean square error criterion was used to assess performances of models. ENET and PCR models provided a significant greater generalisation ability than the DR model ( p = 0.018 , p < 0.001 , p = 0.004 and p < 0.001 for E N E T I , E N E T G , P C R I and P C R G , respectively). Only E N E T G and R F G were significantly more accurate in prediction than DR ( p < 0.001 and p < 0.012 ). In conclusion, ENET achieved greater generalisation and predictive accuracy performances. Thus, building and evaluating models within a generalisation enhancing procedure is a prerequisite for any predictive modelling. [ABSTRACT FROM AUTHOR]

Machine learning can be used to explore the complex multifactorial patterns underlying postsurgical graft detachment after endothelial corneal transplantation surgery and to evaluate the marginal effect of various practice pattern modulations. We included all posterior lamellar keratoplasty procedures recorded in the Dutch Cornea Transplant Registry from 2015 through 2018 and collected the center-specific practice patterns using a questionnaire. All available data regarding the donor, recipient, surgery, and practice pattern, were coded into 91 factors that might be associated with the occurrence of a graft detachment. In this research, we used three machine learning methods; a regularized logistic regression (lasso), classification tree analysis (CTA), and random forest classification (RFC), to select the most predictive subset of variables for graft detachment. A total of 3647 transplants were included in our analysis and the overall prevalence of graft detachment was 9.9%. In an independent test set the area under the curve for the lasso, CTA, and RFC was 0.70, 0.65, and 0.72, respectively. Identified risk factors included: a Descemet membrane endothelial keratoplasty procedure, prior graft failure, and the use of sulfur hexafluoride gas. Factors with a reduced risk included: performing combined procedures, using pre-cut donor tissue, and a pre-operative laser iridotomy. These results can help surgeons to review their practice patterns and generate hypotheses for empirical research regarding the origins of graft detachments. [ABSTRACT FROM AUTHOR]

Machine learning can be used to explore the complex multifactorial patterns underlying postsurgical graft detachment after endothelial corneal transplantation surgery and to evaluate the marginal effect of various practice pattern modulations. We included all posterior lamellar keratoplasty procedures recorded in the Dutch Cornea Transplant Registry from 2015 through 2018 and collected the center-specific practice patterns using a questionnaire. All available data regarding the donor, recipient, surgery, and practice pattern, were coded into 91 factors that might be associated with the occurrence of a graft detachment. In this research, we used three machine learning methods; a regularized logistic regression (lasso), classification tree analysis (CTA), and random forest classification (RFC), to select the most predictive subset of variables for graft detachment. A total of 3647 transplants were included in our analysis and the overall prevalence of graft detachment was 9.9%. In an independent test set the area under the curve for the lasso, CTA, and RFC was 0.70, 0.65, and 0.72, respectively. Identified risk factors included: a Descemet membrane endothelial keratoplasty procedure, prior graft failure, and the use of sulfur hexafluoride gas. Factors with a reduced risk included: performing combined procedures, using pre-cut donor tissue, and a pre-operative laser iridotomy. These results can help surgeons to review their practice patterns and generate hypotheses for empirical research regarding the origins of graft detachments. [ABSTRACT FROM AUTHOR]

Accurate determination of lymph-node (LN) metastases is a prerequisite for high precision radiotherapy. The primary aim is to characterise the performance of PET/CT-based machine-learning classifiers to predict LN-involvement by endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in stage-III NSCLC. Prediction models for LN-positivity based on [18F]FDG-PET/CT features were built using logistic regression and machine-learning models random forest (RF) and multilayer perceptron neural network (MLP) for stage-III NSCLC before radiochemotherapy. A total of 675 LN-stations were sampled in 180 patients. The logistic and RF models identified SUVmax, the short-axis LN-diameter and the echelon of the considered LN among the most important parameters for EBUS-positivity. Adjusting the sensitivity of machine-learning classifiers to that of the expert-rater of 94.5%, MLP (P = 0.0061) and RF models (P = 0.038) showed lower misclassification rates (MCR) than the standard-report, weighting false positives and false negatives equally. Increasing the sensitivity of classifiers from 94.5 to 99.3% resulted in increase of MCR from 13.3/14.5 to 29.8/34.2% for MLP/RF, respectively. PET/CT-based machine-learning classifiers can achieve a high sensitivity (94.5%) to detect EBUS-positive LNs at a low misclassification rate. As the specificity decreases rapidly above that level, a combined test of a PET/CT-based MLP/RF classifier and EBUS-TBNA is recommended for radiation target volume definition. [ABSTRACT FROM AUTHOR]

Machine learning (ML) may be used to predict mortality. We used claims data from one large German insurer to develop and test differently complex ML prediction models, comparing them for their (balanced) accuracy, but also the importance of different predictors, the relevance of the follow-up period before death (i.e. the amount of accumulated data) and the time distance of the data used for prediction and death. A sample of 373,077 insured very old, aged 75 years or above, living in the Northeast of Germany in 2012 was drawn and followed over 6 years. Our outcome was whether an individual died in one of the years of interest (2013–2017) or not; the primary metric was (balanced) accuracy in a hold-out test dataset. From the 86,326 potential variables, we used the 30 most important ones for modeling. We trained a total of 45 model combinations: (1) Three different ML models were used; logistic regression (LR), random forest (RF), extreme gradient boosting (XGB); (2) Different periods of follow-up were employed for training; 1–5 years; (3) Different time distances between data used for prediction and the time of the event (death/survival) were set; 0–4 years. The mortality rate was 9.15% in mean per year. The models showed (balanced) accuracy between 65 and 93%. A longer follow-up period showed limited to no advantage, but models with short time distance from the event were more accurate than models trained on more distant data. RF and XGB were more accurate than LR. For RF and XGB sensitivity and specificity were similar, while for LR sensitivity was significantly lower than specificity. For all three models, the positive-predictive-value was below 62% (and even dropped to below 20% for longer time distances from death), while the negative-predictive-value significantly exceeded 90% for all analyses. The utilization of and costs for emergency transport as well as emergency and any hospital visits as well as the utilization of conventional outpatient care and laboratory services were consistently found most relevant for predicting mortality. All models showed useful accuracies, and more complex models showed advantages. The variables employed for prediction were consistent across models and with medical reasoning. Identifying individuals at risk could assist tailored decision-making and interventions. [ABSTRACT FROM AUTHOR]

Recurrent urinary tract infection (RUTI) can damage renal function and has impact on healthcare costs and patients' quality of life. There were 2 stages for development of prediction models for RUTI. The first stage was a scenario in the clinical visit. The second stage was a scenario after hospitalization for urinary tract infection caused by Escherichia coli. Three machine learning models, logistic regression (LR), decision tree (DT), and random forest (RF) were built for the RUTI prediction. The RF model had higher prediction accuracy than LR and DT (0.700, 0.604, and 0.654 in stage 1, respectively; 0.709, 0.604, and 0.635 in stage 2, respectively). The decision rules constructed by the DT model could provide high classification accuracy (up to 0.92 in stage 1 and 0.94 in stage 2) in certain subgroup patients in different scenarios. In conclusion, this study provided validated machine learning models and RF could provide a better accuracy in predicting the development of single uropathogen (E. coli) RUTI. Both host and bacterial characteristics made important contribution to the development of RUTI in the prediction models in the 2 clinical scenarios, respectively. Based on the results, physicians could take action to prevent the development of RUTI. [ABSTRACT FROM AUTHOR]

Cuffless blood pressure measurement enables unobtrusive and continuous monitoring that can be integrated with wearable devices to extend healthcare to non-hospital settings. Most of the current research has focused on the estimation of blood pressure based on pulse transit time or pulse arrival time using ECG or peripheral cardiac pulse signals as proximal time references. This study proposed the use of a phonocardiogram (PCG) and ballistocardiogram (BCG), two signals detected noninvasively, to estimate systolic blood pressure (SBP). For this, the PCG and the BCG were simultaneously measured in 21 volunteers in the rest, activity, and post-activity conditions. Different features were considered based on the relationships between these signals. The intervals between S1 and S2 of the PCG and the I, J, and K waves of the BCG were statistically analyzed. The IJ and JK slopes were also estimated as additional features to train the machine-learning algorithm. The intervals S1-J, S1-K, S1-I, J-S2, and I-S2 were negatively correlated with changes in SBP (p-val < 0.01). The features were used as explanatory variables for a regressor based on the Random Forest. It was possible to estimate the systolic blood pressure with a mean error of 3.3 mmHg with a standard deviation of ± 5 mmHg. Therefore, we foresee that this proposal has potential applications for wearable devices that use low-cost embedded systems. [ABSTRACT FROM AUTHOR]

Internet of Things (IoT) 's devices are ubiquitous and operate in a heterogonous environment with potential security breaches. IoT Operating Systems (IoT OSs) are the backbone software for running such devices. If IoT OSs are vulnerable to security breaches, higher-level security measures may not help. This paper aims to use Machine Learning (ML) to create a tool called iDetect for detecting vulnerabilities in C/C++ source code of IoT OSs. The source code for 16 releases of IoT OSs (RIOT, Contiki, FreeRTOS, Amazon FreeRTOS) and the Software Assurance Reference Dataset (SARD) were used to create a labeled dataset of vulnerable and benign code with the reference being the Common Weakness Enumeration (CWE) vulnerabilities present in IoT OSs. Studies showed that only a subset of CWEs is present in the C/C++ source code of low-end IoT OSs.The labeled dataset was used to train three ML models for vulnerability detection: Random Forest (RF), Convolutional Neural Network (CNN), and Recurrent Neural Network (RNN). The three models were used independently and RF; compared to CNN and RNN, gave the highest accuracy during the testing phase for binary and multiclass classification. RF was chosen as iDetect's ML classifier. Further evaluation was done on an unseen dataset of 322 code snippets taken from TinyOS. iDetect achieved a macro-averaged F1 score (mF1) of 98.5% and weighted-average F1 score (wF1) of 98% for multiclass classification, F1 score (F1) of 97.8% for binary classification, and superior results compared to all three Static Analysis Tools (SATs) used to collect the training dataset. [ABSTRACT FROM AUTHOR]

Sigmoid colon cancer often has an unsatisfactory prognosis. This study explored the effect of tumor deposits (TDs) on survival, and whether their presence/absence influence individualized treatment. Data of postoperative patients with sigmoid colon cancer were extracted from the Surveillance, Epidemiology, and End Results database. Overall survival (OS) was calculated using the Kaplan–Meier method and prognostic factors were identified using Cox regression analysis and random forest (RF). The nomogram's discrimination performance was evaluated using a concordance index (C-index), integrated discrimination improvement (IDI), calibration curves, and decision-curve analysis. The N1c group showed a worse prognosis than the N0 group. For N1c patients, a combination of surgery and chemotherapy prolonged survival, compared to surgery alone; however, the chemotherapy-surgery combination did not affect the OS of patients younger than 70 years, in stage T1–2, and/or of black race. Multivariable analysis and RF presented Age, T stage, and N stage were the most important predictors for OS. The novel nomogram had superiority to the TNM staging system with improved C-index and IDI, as well as good consistency and higher clinical benefit. TDs are associated with poor survival from sigmoid colon cancer, and considering TDs can inform the formulation of individual treatment regimens. The nomogram shows satisfactory prediction ability for OS. [ABSTRACT FROM AUTHOR]

Preventing unplanned hospitalisations, including readmissions and re-presentations to the emergency department, is an important strategy for addressing the growing demand for hospital care. Significant successes have been reported from interventions put in place by hospitals to reduce their incidence. However, there is limited use of data-driven algorithms in hospital services to identify patients for enrolment into these intervention programs. Here we present the results of a study aiming to develop algorithms deployable at scale as part of a state government's initiative to address rehospitalizations and which fills several gaps identified in the state-of-the-art literature. To the best of our knowledge, our study involves the largest-ever sample size for developing risk models. Logistic regression, random forests and gradient boosted techniques were explored as model candidates and validated retrospectively on five years of data from 27 hospitals in Queensland, Australia. The models used a range of predictor variables sourced from state-wide Emergency Department(ED), inpatient, hospital-dispensed medications and hospital-requested pathology databases. The investigation leads to several findings: (i) the advantage of looking at a longer patient data history, (ii) ED and inpatient datasets alone can provide useful information for predicting hospitalisation risk and the addition of medications and pathology test results leads to trivial performance improvements, (iii) predicting readmissions to the hospital was slightly easier than predicting re-presentations to ED after an inpatient stay, which was slightly easier again than predicting re-presentations to ED after an EDstay, (iv) a gradient boosted approach (XGBoost) was systematically the most powerful modelling approach across various tests. [ABSTRACT FROM AUTHOR]

An important structuring feature of a soccer match is the in-game status, whether a match is interrupted or in play. This is necessary to calculate performance indicators relative to the effective playing time or to find standard situations, ball actions, and other tactical structures in spatiotemporal data. Our study explores the extent to which the in-game status can be determined using time-continuous player positions. Therefore, to determine the in-game status we tested four established machine learning methods: logistic regression, decision trees, random forests, and AdaBoost. The models were trained and evaluated using spatiotemporal data and manually annotated in-game status of 102 matches in the German Bundesliga. Results show up to 92% accuracy in predicting the in-game status in previously unknown matches on frame level. The best performing method, AdaBoost, shows 81% precision for detecting stoppages (longer than 2 s). The absolute time shift error at the start was ≤ 2 s for 77% and 81% at the end for all correctly predicted stoppages. The mean error of the in-game total distance covered per player per match using the AdaBoost in-game status prediction was − 102 ± 273 m, which is 1.3% of the mean value of this performance indicator (7939 m). Conclusively, the prediction quality of our model is high enough to provide merit for performance diagnostics when teams have access to player positions (e.g., from GPS/LPM systems) but no human-annotated in-game status and/or ball position data, such as in amateur or youth soccer. [ABSTRACT FROM AUTHOR]

Independent coffee shops are the alternative workplaces for people working remotely from traditional offices but are not concerned about their indoor air quality (IAQ). This study aimed to rank the environmental factors in affecting the IAQ by Random Forests (RFs) models. The indoor environments and human activities of participated independent coffee shops were observed and recorded for 3 consecutive days including weekdays and weekend during the business hours. The multi-sized particulate matter (PM), particle-bound polycyclic aromatic hydrocarbons (p-PAHs), total volatile organic compounds (TVOCs), CO, CO2, temperature and relative humidity were monitored. RFs models ranked the environmental factors. More than 20% of the 15-min average concentrations of PM10, PM2.5, and CO2 exceeded the World Health Organization guidelines. Occupant density affected TVOCs, p-PAHs and CO2 concentrations directly. Tobacco smoking dominated PM10, PM2.5, TVOCs and p-PAHs concentrations mostly. CO concentration was affected by roasting bean first and tobacco smoking secondly. The non-linear relationships between temperature and these pollutants illustrated the relative low concentrations happened at temperature between 22 and 24 °C. Tobacco smoking, roasting beans and occupant density are the observable activities to alert the IAQ change. Decreasing CO2 and optimizing the room temperature could also be the surrogate parameters to assure the IAQ. [ABSTRACT FROM AUTHOR]

Leaf area index (LAI) is a fundamental indicator of crop growth status, timely and non-destructive estimation of LAI is of significant importance for precision agriculture. In this study, a multi-rotor UAV platform equipped with CMOS image sensors was used to capture maize canopy information, simultaneously, a total of 264 ground‐measured LAI data were collected during a 2-year field experiment. Linear regression (LR), backpropagation neural network (BPNN), and random forest (RF) algorithms were used to establish LAI estimation models, and their performances were evaluated through 500 repetitions of random sub-sampling, training, and testing. The results showed that RGB-based VIs derived from UAV digital images were strongly related to LAI, and the grain-filling stage (GS) of maize was identified as the optimal period for LAI estimation. The RF model performed best at both whole period and individual growth stages, with the highest R2 (0.71–0.88) and the lowest RMSE (0.12–0.25) on test datasets, followed by the BPNN model and LR models. In addition, a smaller 5–95% interval range of R2 and RMSE was observed in the RF model, which indicated that the RF model has good generalization ability and is able to produce reliable estimation results. [ABSTRACT FROM AUTHOR]

Identifying disease-associated susceptibility loci is one of the most pressing and crucial challenges in modeling complex diseases. Existing approaches to biomarker discovery are subject to several limitations including underpowered detection, neglect for variant interactions, and restrictive dependence on prior biological knowledge. Addressing these challenges necessitates more ingenious ways of approaching the "missing heritability" problem. This study aims to discover disease-associated susceptibility loci by augmenting previous genome-wide association study (GWAS) using the integration of random forest and cluster analysis. The proposed integrated framework is applied to a hepatitis B virus surface antigen (HBsAg) seroclearance GWAS data. Multiple cluster analyses were performed on (1) single nucleotide polymorphisms (SNPs) considered significant by GWAS and (2) SNPs with the highest feature importance scores obtained using random forest. The resulting SNP-sets from the cluster analyses were subsequently tested for trait-association. Three susceptibility loci possibly associated with HBsAg seroclearance were identified: (1) SNP rs2399971, (2) gene LINC00578, and (3) locus 11p15. SNP rs2399971 is a biomarker reported in the literature to be significantly associated with HBsAg seroclearance in patients who had received antiviral treatment. The latter two loci are linked with diseases influenced by the presence of hepatitis B virus infection. These findings demonstrate the potential of the proposed integrated framework in identifying disease-associated susceptibility loci. With further validation, results herein could aid in better understanding complex disease etiologies and provide inputs for a more advanced disease risk assessment for patients. [ABSTRACT FROM AUTHOR]

Individual daily travel activities (e.g., work, eating) are identified with various machine learning models (e.g., Bayesian Network, Random Forest) for understanding people's frequent travel purposes. However, labor-intensive engineering work is often required to extract effective features. Additionally, features and models are mostly calibrated for individual trajectories with regular daily travel routines and patterns, and therefore suffer from poor generalizability when applied to new trajectories with more irregular patterns. Meanwhile, most existing models cannot extract features to explicitly represent regular travel activity sequences. Therefore, this paper proposes a graph-based representation of spatiotemporal trajectories and point-of-interest (POI) data for travel activity type identification, defined as Gstp2Vec. Specifically, a weighted directed graph is constructed by connecting regular activity areas (i.e., zones) detected via clustering individual daily travel trajectories as graph nodes, with edges denoting trips between pairs of zones. Statistics of trajectories (e.g., visit frequency, activity duration) and POI distributions (e.g., percentage of restaurants) at each activity zone are encoded as node features. Next, trip frequency, average trip duration, and average trip distance are encoded as edge weights. Then a series of feedforward neural networks are trained to generate low-dimensional embeddings for activity nodes through sampling and aggregating spatiotemporal and POI features from their multihop neighborhoods. Activity type labels collected via travel surveys are used as ground truth for backpropagation. The experiment results with real-world GPS trajectories show that Gstp2Vec significantly reduces feature engineering efforts by automatically learning feature embeddings from raw trajectories with minimal prepossessing efforts. It not only enhances model generalizability to receive higher identification accuracy on test individual trajectories with diverse travel patterns, but also obtains better efficiency and robustness. In particular, our identification of the most common daily travel activities (e.g., Dwelling and Work) for people with diverse travel patterns outperforms state-of-the-art classification models. [ABSTRACT FROM AUTHOR]

The original article can be found online at https://doi.org/10.1038/s41598-022-11422-2. Correction to: I Scientific Reports i https://doi.org/10.1038/s41598-022-11422-2, published online 27 May 2022 In the original version of this Article, Figures 14, 15 and 16 were incorrect. [Extracted from the article]

Random forests are a popular type of machine learning model, which are relatively robust to overfitting, unlike some other machine learning models, and adequately capture non-linear relationships between an outcome of interest and multiple independent variables. There are relatively few adjustable hyperparameters in the standard random forest models, among them the minimum size of the terminal nodes on each tree. The usual stopping rule, as proposed by Breiman, stops tree expansion by limiting the size of the parent nodes, so that a node cannot be split if it has less than a specified number of observations. Recently an alternative stopping criterion has been proposed, stopping tree expansion so that all terminal nodes have at least a minimum number of observations. The present paper proposes three generalisations of this idea, limiting the growth in regression random forests, based on the variance, range, or inter-centile range. The new approaches are applied to diabetes data obtained from the National Health and Nutrition Examination Survey and four other datasets (Tasmanian Abalone data, Boston Housing crime rate data, Los Angeles ozone concentration data, MIT servo data). Empirical analysis presented herein demonstrate that the new stopping rules yield competitive mean square prediction error to standard random forest models. In general, use of the intercentile range statistic to control tree expansion yields much less variation in mean square prediction error, and mean square prediction error is also closer to the optimal. The Fortran code developed is provided in the Supplementary Material. [ABSTRACT FROM AUTHOR]

Suppose we are given a system of coupled oscillators on an unknown graph along with the trajectory of the system during some period. Can we predict whether the system will eventually synchronize? Even with a known underlying graph structure, this is an important yet analytically intractable question in general. In this work, we take an alternative approach to the synchronization prediction problem by viewing it as a classification problem based on the fact that any given system will eventually synchronize or converge to a non-synchronizing limit cycle. By only using some basic statistics of the underlying graphs such as edge density and diameter, our method can achieve perfect accuracy when there is a significant difference in the topology of the underlying graphs between the synchronizing and the non-synchronizing examples. However, in the problem setting where these graph statistics cannot distinguish the two classes very well (e.g., when the graphs are generated from the same random graph model), we find that pairing a few iterations of the initial dynamics along with the graph statistics as the input to our classification algorithms can lead to significant improvement in accuracy; far exceeding what is known by the classical oscillator theory. More surprisingly, we find that in almost all such settings, dropping out the basic graph statistics and training our algorithms with only initial dynamics achieves nearly the same accuracy. We demonstrate our method on three models of continuous and discrete coupled oscillators—the Kuramoto model, Firefly Cellular Automata, and Greenberg-Hastings model. Finally, we also propose an "ensemble prediction" algorithm that successfully scales our method to large graphs by training on dynamics observed from multiple random subgraphs. [ABSTRACT FROM AUTHOR]

There has been a growing attention on using machine learning (ML) in pharmacovigilance. This study aimed to investigate the utility of supervised ML algorithms on timely detection of safety signals in the Korea Adverse Event Reporting System (KAERS), using infliximab as a case drug, between 2009 and 2018. Input data set for ML training was constructed based on the drug label information and spontaneous reports in the KAERS. Gold standard dataset containing known AEs was randomly divided into the training and test sets. Two supervised ML algorithms (gradient boosting machine [GBM], random forest [RF]) were fitted with hyperparameters tuned on the training set by using a fivefold validation. Then, we stratified the KAERS data by calendar year to create 10 cumulative yearly datasets, in which ML algorithms were applied to detect five pre-specified AEs of infliximab identified during post-marketing surveillance. Four AEs were detected by both GBM and RF in the first year they appeared in the KAERS and earlier than they were updated in the drug label of infliximab. We further applied our models to data retrieved from the US Food and Drug Administration Adverse Event Reporting System repository and found that they outperformed existing disproportionality methods. Both GBM and RF demonstrated reliable performance in detecting early safety signals and showed promise for applying such approaches to pharmacovigilance. [ABSTRACT FROM AUTHOR]

Seaweeds are an important source of nutrients and bioactive compounds and have a high potential as health boosters in aquaculture. This study evaluated the effect of dietary inclusion of Gracilaria gracilis biomass or its extract on the European seabass (Dicentrarchus labrax) gut microbial community. Juvenile fish were fed a commercial-like diet with 2.5% or 5% seaweed biomass or 0.35% seaweed extract for 47 days. The gut microbiome was assessed by 16S rRNA amplicon sequencing, and its diversity was not altered by the seaweed supplementation. However, a reduction in Proteobacteria abundance was observed. Random forest analysis highlighted the genera Photobacterium, Staphylococcus, Acinetobacter, Micrococcus and Sphingomonas, and their abundances were reduced when fish were fed diets with algae. SparCC correlation network analysis suggested several mutualistic and other antagonistic relationships that could be related to the predicted altered functions. These pathways were mainly related to the metabolism and biosynthesis of protective compounds such as ectoine and were upregulated in fish fed diets supplemented with algae. This study shows the beneficial potential of Gracilaria as a functional ingredient through the modulation of the complex microbial network towards fish health improvement. [ABSTRACT FROM AUTHOR]

Identification of soft rot disease in napa cabbage, an essential ingredient of kimchi, is challenging at the industrial scale. Therefore, nondestructive imaging techniques are necessary. Here, we investigated the potential of hyperspectral imaging (HSI) processing in the near-infrared region (900–1700 nm) for classifying napa cabbage quality using nondestructive measurements. We determined the microbiological and physicochemical qualitative properties of napa cabbage for intercomparison of HSI information, extracted HSI characteristics from hyperspectral images to predict and classify freshness, and established a novel approach for classifying healthy and rotten napa cabbage. The second derivative Savitzky–Golay method for data preprocessing was implemented, followed by wavelength selection using variable importance in projection scores. For multivariate data of the classification models, partial least square discriminant analysis (PLS-DA), support vector machine (SVM), and random forests were used for predicting cabbage conditions. The SVM model accurately distinguished the cabbage exhibiting soft rot disease symptoms from the healthy cabbage. This study presents the potential of HSI systems for separating soft rot disease-infected napa cabbages from healthy napa cabbages using the SVM model, especially under the most effective wavelengths (970, 980, 1180, 1070, 1120, and 978 nm), prior to processing. These results are applicable to industrial multispectral images. [ABSTRACT FROM AUTHOR]

Identification of soft rot disease in napa cabbage, an essential ingredient of kimchi, is challenging at the industrial scale. Therefore, nondestructive imaging techniques are necessary. Here, we investigated the potential of hyperspectral imaging (HSI) processing in the near-infrared region (900–1700 nm) for classifying napa cabbage quality using nondestructive measurements. We determined the microbiological and physicochemical qualitative properties of napa cabbage for intercomparison of HSI information, extracted HSI characteristics from hyperspectral images to predict and classify freshness, and established a novel approach for classifying healthy and rotten napa cabbage. The second derivative Savitzky–Golay method for data preprocessing was implemented, followed by wavelength selection using variable importance in projection scores. For multivariate data of the classification models, partial least square discriminant analysis (PLS-DA), support vector machine (SVM), and random forests were used for predicting cabbage conditions. The SVM model accurately distinguished the cabbage exhibiting soft rot disease symptoms from the healthy cabbage. This study presents the potential of HSI systems for separating soft rot disease-infected napa cabbages from healthy napa cabbages using the SVM model, especially under the most effective wavelengths (970, 980, 1180, 1070, 1120, and 978 nm), prior to processing. These results are applicable to industrial multispectral images. [ABSTRACT FROM AUTHOR]

Walkability is an important measure with strong ties to our health. However, there are existing gaps in the literature. Our previous work proposed new approaches to address existing limitations. This paper explores new ways of applying transferability using transfer-learning. Road networks, POIs, and road-related characteristics grow/change over time. Moreover, calculating walkability for all locations in all cities is very time-consuming. Transferability enables reuse of already-learned knowledge for continued learning, reduce training time, resource consumption, training labels and improve prediction accuracy. We propose ALF-Score++, that reuses trained models to generate transferable models capable of predicting walkability score for cities not seen in the process. We trained transfer-learned models for St. John's NL and Montréal QC and used them to predict walkability scores for Kingston ON and Vancouver BC. MAE error of 13.87 units (ranging 0–100) was achieved for transfer-learning using MLP and 4.56 units for direct-training (random forest) on personalized clusters. [ABSTRACT FROM AUTHOR]

Machine learning methods trained on cancer cell line panels are intensively studied for the prediction of optimal anti-cancer therapies. While classification approaches distinguish effective from ineffective drugs, regression approaches aim to quantify the degree of drug effectiveness. However, the high specificity of most anti-cancer drugs induces a skewed distribution of drug response values in favor of the more drug-resistant cell lines, negatively affecting the classification performance (class imbalance) and regression performance (regression imbalance) for the sensitive cell lines. Here, we present a novel approach called SimultAneoUs Regression and classificatiON Random Forests (SAURON-RF) based on the idea of performing a joint regression and classification analysis. We demonstrate that SAURON-RF improves the classification and regression performance for the sensitive cell lines at the expense of a moderate loss for the resistant ones. Furthermore, our results show that simultaneous classification and regression can be superior to regression or classification alone. [ABSTRACT FROM AUTHOR]

Peripheral artery disease (PAD) is a common cardiovascular disorder that is frequently underdiagnosed, which can lead to poorer outcomes due to lower rates of medical optimization. We aimed to develop an automated tool to identify undiagnosed PAD and evaluate physician acceptance of a dashboard representation of risk assessment. Data were derived from electronic health records (EHR). We developed and compared traditional risk score models to novel machine learning models. For usability testing, primary and specialty care physicians were recruited and interviewed until thematic saturation. Data from 3168 patients with PAD and 16,863 controls were utilized. Results showed a deep learning model that utilized time engineered features outperformed random forest and traditional logistic regression models (average AUCs 0.96, 0.91 and 0.81, respectively), P < 0.0001. Of interviewed physicians, 75% were receptive to an EHR-based automated PAD model. Feedback emphasized workflow optimization, including integrating risk assessments directly into the EHR, using dashboard designs that minimize clicks, and providing risk assessments for clinically complex patients. In conclusion, we demonstrate that EHR-based machine learning models can accurately detect risk of PAD and that physicians are receptive to automated risk detection for PAD. Future research aims to prospectively validate model performance and impact on patient outcomes. [ABSTRACT FROM AUTHOR]

In this study, a machine learning model for the precise manufacturing of green cementitious composites modified with granite powder sourced from quarry waste was designed. For this purpose, decision tree, random forest and AdaBoost ensemble models were used and compared. A database was created containing 216 sets of data based on an experimental study. The database consists of parameters such as the percentage of cement substituted with granite powder, time of testing and curing conditions. It was shown that this method for designing green cementitious composite mixes, in terms of predicting compressive strength using ensemble models and only three input parameters, can be more accurate and much more precise than the conventional approach. Moreover, to the best of the authors' knowledge, artificial intelligence has been one of the most effective and precise methods used in the design and manufacturing industry in recent decades. The simplicity of this method makes it more suitable for construction practice due to the ease of evaluating the input variables. As the push towards decreasing carbon emissions increases, a method for designing green cementitious composites without producing waste that is more precise than traditional tests performed in a laboratory is essential. [ABSTRACT FROM AUTHOR]

Wetland vegetation classification using deep learning algorithm and unmanned aerial vehicle (UAV) images have attracted increased attentions. However, there exist several challenges in mapping karst wetland vegetation due to its fragmentation, intersection, and high heterogeneity of vegetation patches. This study proposed a novel approach to classify karst vegetation in Huixian National Wetland Park, the largest karst wetland in China by fusing single-class SegNet classification using the maximum probability algorithm. A new optimized post-classification algorithm was developed to eliminate the stitching traces caused by SegNet model prediction. This paper evaluated the effect of multi-class and fusion of multiple single-class SegNet models with different EPOCH values on mapping karst vegetation using UAV images. Finally, this paper carried out a comparison of classification accuracies between object-based Random Forest (RF) and fusion of single-class SegNet models. The specific conclusions of this paper include the followings: (1) fusion of four single-class SegNet models produced better classification for karst wetland vegetation than multi-class SegNet model, and achieved the highest overall accuracy of 87.34%; (2) the optimized post-classification algorithm improved classification accuracy of SegNet model by eliminating splicing traces; (3) classification performance of single-class SegNet model outperformed multi-class SegNet model, and improved classification accuracy (F1-Score) ranging from 10 to 25%; (4) Fusion of single-class SegNet models and object-based RF classifier both produced good classifications for karst wetland vegetation, and achieved over 87% overall accuracy. [ABSTRACT FROM AUTHOR]

Evaporation is the primary aspect causing water loss in the hydrological cycle; therefore, water loss must be precisely measured. Evaporation is an intricate nonlinear process occurring as a result of several climatic aspects. The purpose of this research is to assess the feasibility of using Random Forest (RF) and two deep learning techniques, namely convolutional neural network (CNN), and deep neural network (DNN) to accurately estimate monthly pan evaporation rates. Month-based weather data gathered from four Malaysian weather stations during the 2000–2019 timeframe was used to train and evaluate the models. Several input attributes (predictor variables) were investigated to select the most suitable variables for machine learning models. Every approach was tested with several models, each with a different set of model aspects and input parameter combinations. The formulated ML approaches were benchmarked against two commonly used empirical methods: Stephens & Stewart and Thornthwaite. Model outcomes were assessed using standard statistical measures to determine their effectiveness in predicting evaporation. The results indicated that the three ML models developed in the study performed better than empirical models and could significantly improve the precision of monthly Ep estimates even with the identical input sets. The performance assessment metrics also show that the formulated CNN approach was acceptable for modelling monthly water loss due to evaporation with a higher degree of accuracy than other ML frameworks explored in this study. In addition, the CNN framework outperformed other AI techniques evaluated for the same areas using identical data inputs. The investigation's findings in relation to the various performance criteria show that the proposed CNN model is capable of capturing the highly non-linearity of evaporation and could be regarded as an effective tool to predict evaporation. [ABSTRACT FROM AUTHOR]

HIV treatment programs face challenges in identifying patients at risk for loss-to-follow-up and uncontrolled viremia. We applied predictive machine learning algorithms to anonymised, patient-level HIV programmatic data from two districts in South Africa, 2016–2018. We developed patient risk scores for two outcomes: (1) visit attendance ≤ 28 days of the next scheduled clinic visit and (2) suppression of the next HIV viral load (VL). Demographic, clinical, behavioral and laboratory data were investigated in multiple models as predictor variables of attending the next scheduled visit and VL results at the next test. Three classification algorithms (logistical regression, random forest and AdaBoost) were evaluated for building predictive models. Data were randomly sampled on a 70/30 split into a training and test set. The training set included a balanced set of positive and negative examples from which the classification algorithm could learn. The predictor variable data from the unseen test set were given to the model, and each predicted outcome was scored against known outcomes. Finally, we estimated performance metrics for each model in terms of sensitivity, specificity, positive and negative predictive value and area under the curve (AUC). In total, 445,636 patients were included in the retention model and 363,977 in the VL model. The predictive metric (AUC) ranged from 0.69 for attendance at the next scheduled visit to 0.76 for VL suppression, suggesting that the model correctly classified whether a scheduled visit would be attended in 2 of 3 patients and whether the VL result at the next test would be suppressed in approximately 3 of 4 patients. Variables that were important predictors of both outcomes included prior late visits, number of prior VL tests, time since their last visit, number of visits on their current regimen, age, and treatment duration. For retention, the number of visits at the current facility and the details of the next appointment date were also predictors, while for VL suppression, other predictors included the range of the previous VL value. Machine learning can identify HIV patients at risk for disengagement and unsuppressed VL. Predictive modeling can improve the targeting of interventions through differentiated models of care before patients disengage from treatment programmes, increasing cost-effectiveness and improving patient outcomes. [ABSTRACT FROM AUTHOR]

Melia azedarach L. is an important economic tree widely distributed in tropical and subtropical regions of China and some other countries. However, it is unclear how the species' suitable habitat will respond to future climate changes. We aimed to select the most accurate one among seven data mining models to predict the current and future suitable habitats for M. azedarach in China. These models include: maximum entropy (MaxEnt), support vector machine (SVM), generalized linear model (GLM), random forest (RF), naive bayesian model (NBM), extreme gradient boosting (XGBoost), and gradient boosting machine (GBM). A total of 906 M. azedarach locations were identified, and sixteen climate predictors were used for model building. The models' validity was assessed using three measures (Area Under the Curves (AUC), kappa, and overall accuracy (OA)). We found that the RF provided the most outstanding performance in prediction power and generalization capacity. The top climate factors affecting the species' suitable habitats were mean coldest month temperature (MCMT), followed by the number of frost-free days (NFFD), degree-days above 18 °C (DD > 18), temperature difference between MWMT and MCMT, or continentality (TD), mean annual precipitation (MAP), and degree-days below 18 °C (DD < 18). We projected that future suitable habitat of this species would increase under both the RCP4.5 and RCP8.5 scenarios for the 2011–2040 (2020s), 2041–2070 (2050s), and 2071–2100 (2080s). Our findings are expected to assist in better understanding the impact of climate change on the species and provide scientific basis for its planting and conservation. [ABSTRACT FROM AUTHOR]

Precision medicine applies machine learning methods to estimate the personalized optimal treatment decision based on individual information, such as genetic data and medical history. The main purpose of self obesity management is to develop a personalized optimal life plan that is easy to implement and adhere to, thereby reducing the incidence of obesity and obesity-related diseases. The methodology comprises three components. First, we apply catboost, random forest and lasso covariance test to evaluate the importance of individual features in forecasting body mass index. Second, we apply metaalgorithms to estimate the personalized optimal decision on alcohol, vegetable, high caloric food and daily water intake respectively for each individual. Third, we propose new metaalgorithms named SX and SXwint learners to compute the personalized optimal decision and compare their performances with other prevailing metalearners. We find that people who receive individualized optimal treatment options not only have lower obesity levels than others, but also have lower obesity levels than those who receive 'one-for-all' treatment options. In conclusion, all metaalgorithms are effective at estimating the personalized optimal decision, where SXwint learner shows the best performance on daily water intake. [ABSTRACT FROM AUTHOR]

Accurate prediction of a newborn's birth weight (BW) is a crucial determinant to evaluate the newborn's health and safety. Infants with low BW (LBW) are at a higher risk of serious short- and long-term health outcomes. Over the past decade, machine learning (ML) techniques have shown a successful breakthrough in the field of medical diagnostics. Various automated systems have been proposed that use maternal features for LBW prediction. However, each proposed system uses different maternal features for LBW classification and estimation. Therefore, this paper provides a detailed setup for BW estimation and LBW classification. Multiple subsets of features were combined to perform predictions with and without feature selection techniques. Furthermore, the synthetic minority oversampling technique was employed to oversample the minority class. The performance of 30 ML algorithms was evaluated for both infant BW estimation and LBW classification. Experiments were performed on a self-created dataset with 88 features. The dataset was obtained from 821 women from three hospitals in the United Arab Emirates. Different performance metrics, such as mean absolute error and mean absolute percent error, were used for BW estimation. Accuracy, precision, recall, F-scores, and confusion matrices were used for LBW classification. Extensive experiments performed using five-folds cross validation show that the best weight estimation was obtained using Random Forest algorithm with mean absolute error of 294.53 g while the best classification performance was obtained using Logistic Regression with SMOTE oversampling techniques that achieved accuracy, precision, recall and F1 score of 90.24%, 87.6%, 90.2% and 0.89, respectively. The results also suggest that features such as diabetes, hypertension, and gestational age, play a vital role in LBW classification. [ABSTRACT FROM AUTHOR]

We aimed to investigate the status of falls and to identify important risk factors for falls in persons with type 2 diabetes (T2D) including the non-elderly. Participants were 316 persons with T2D who were assessed for medical history, laboratory data and physical capabilities during hospitalization and given a questionnaire on falls one year after discharge. Two different statistical models, logistic regression and random forest classifier, were used to identify the important predictors of falls. The response rate to the survey was 72%; of the 226 respondents, there were 129 males and 97 females (median age 62 years). The fall rate during the first year after discharge was 19%. Logistic regression revealed that knee extension strength, fasting C-peptide (F-CPR) level and dorsiflexion strength were independent predictors of falls. The random forest classifier placed grip strength, F-CPR, knee extension strength, dorsiflexion strength and proliferative diabetic retinopathy among the 5 most important variables for falls. Lower extremity muscle weakness, elevated F-CPR levels and reduced grip strength were shown to be important risk factors for falls in T2D. Analysis by random forest can identify new risk factors for falls in addition to logistic regression. [ABSTRACT FROM AUTHOR]

Glass-forming ability (GFA) of bulk metallic glasses (BMGs) is a determinant parameter which has been significantly studied. GFA improvements could be achieved through trial-and-error experiments, as a tedious work, or by using developed predicting tools. Machine-Learning (ML) has been used as a promising method to predict the properties of BMGs by removing the barriers in the way of its alloy design. This article aims to develop a ML-based method for predicting the maximum critical diameter (Dmax) of BMGs as a factor of their glass-forming ability. The main result is that the random forest method can be used as a sustainable model (R2 = 92%) for predicting glass-forming ability. Also, adding characteristic temperatures to the model will increase the accuracy and efficiency of the developed model. Comparing the measured and predicted values of Dmax for a set of newly developed BMGs indicated that the model is reliable and can be truly used for predicting the GFA of BMGs. [ABSTRACT FROM AUTHOR]

Coronary artery disease (CAD) is a prevalent disease with high morbidity and mortality rates. Invasive coronary angiography is the reference standard for diagnosing CAD but is costly and associated with risks. Noninvasive imaging like cardiac magnetic resonance (CMR) facilitates CAD assessment and can serve as a gatekeeper to downstream invasive testing. Machine learning methods are increasingly applied for automated interpretation of imaging and other clinical results for medical diagnosis. In this study, we proposed a novel CAD detection method based on CMR images by utilizing the feature extraction ability of deep neural networks and combining the features with the aid of a random forest for the very first time. It is necessary to convert image data to numeric features so that they can be used in the nodes of the decision trees. To this end, the predictions of multiple stand-alone convolutional neural networks (CNNs) were considered as input features for the decision trees. The capability of CNNs in representing image data renders our method a generic classification approach applicable to any image dataset. We named our method RF-CNN-F, which stands for Random Forest with CNN Features. We conducted experiments on a large CMR dataset that we have collected and made publicly accessible. Our method achieved excellent accuracy (99.18%) using Adam optimizer compared to a stand-alone CNN trained using fivefold cross validation (93.92%) tested on the same dataset. [ABSTRACT FROM AUTHOR]

Airplanes have always been one of the first choices for people to travel because of their convenience and safety. However, due to the outbreak of the new coronavirus epidemic in 2020, the civil aviation industry of various countries in the world has encountered severe challenges. Predicting aircraft passenger satisfaction and excavating the main influencing factors can help airlines improve their services and gain advantages in difficult situations and competition. This paper proposes a RF-RFE-Logistic feature selection model to extract the influencing factors of passenger satisfaction. First, preliminary feature selection is performed using recursive feature elimination based on random forest (RF-RFE). Second, based on different classification models, KNN, logistic regression, random forest, Gaussian Naive Bayes, and BP neural network, the classification performance of the models before and after feature selection is compared, and the prediction model with the best classification performance is selected. Finally, based on the RF-RFE feature selection, combined with the logistic model, the factors affecting customer satisfaction are further extracted. The experimental results show that the RF-RFE model selects a feature subset containing 17 variables. In the classification prediction model, the random forest after RF-RFE feature selection shows the best classification performance. Finally, combined with the four important variables extracted by RF-RFE and logistic regression, further discussion is carried out, and suggestions are given for airlines to improve passenger satisfaction. [ABSTRACT FROM AUTHOR]

During spaceflight, astronauts face a unique set of stressors, including microgravity, isolation, and confinement, as well as environmental and operational hazards. These factors can negatively impact sleep, alertness, and neurobehavioral performance, all of which are critical to mission success. In this paper, we predict neurobehavioral performance over the course of a 6-month mission aboard the International Space Station (ISS), using ISS environmental data as well as self-reported and cognitive data collected longitudinally from 24 astronauts. Neurobehavioral performance was repeatedly assessed via a 3-min Psychomotor Vigilance Test (PVT-B) that is highly sensitive to the effects of sleep deprivation. To relate PVT-B performance to time-varying and discordantly-measured environmental, operational, and psychological covariates, we propose an ensemble prediction model comprising of linear mixed effects, random forest, and functional concurrent models. An extensive cross-validation procedure reveals that this ensemble outperforms any one of its components alone. We also identify the most important predictors of PVT-B performance, which include an individual's previous PVT-B performance, reported fatigue and stress, and temperature and radiation dose. This method is broadly applicable to settings where the main goal is accurate, individualized prediction of human behavior involving a mixture of person-level traits and irregularly measured time series. [ABSTRACT FROM AUTHOR]

The timely identification of cohort participants at higher risk for attrition is important to earlier interventions and efficient use of research resources. Machine learning may have advantages over the conventional approaches to improve discrimination by analysing complex interactions among predictors. We developed predictive models of attrition applying a conventional regression model and different machine learning methods. A total of 542 very preterm (< 32 gestational weeks) infants born in Portugal as part of the European Effective Perinatal Intensive Care in Europe (EPICE) cohort were included. We tested a model with a fixed number of predictors (Baseline) and a second with a dynamic number of variables added from each follow-up (Incremental). Eight classification methods were applied: AdaBoost, Artificial Neural Networks, Functional Trees, J48, J48Consolidated, K-Nearest Neighbours, Random Forest and Logistic Regression. Performance was compared using AUC- PR (Area Under the Curve—Precision Recall), Accuracy, Sensitivity and F-measure. Attrition at the four follow-ups were, respectively: 16%, 25%, 13% and 17%. Both models demonstrated good predictive performance, AUC-PR ranging between 69 and 94.1 in Baseline and from 72.5 to 97.1 in Incremental model. Of the whole set of methods, Random Forest presented the best performance at all follow-ups [AUC-PR1: 94.1 (2.0); AUC-PR2: 91.2 (1.2); AUC-PR3: 97.1 (1.0); AUC-PR4: 96.5 (1.7)]. Logistic Regression performed well below Random Forest. The top-ranked predictors were common for both models in all follow-ups: birthweight, gestational age, maternal age, and length of hospital stay. Random Forest presented the highest capacity for prediction and provided interpretable predictors. Researchers involved in cohorts can benefit from our robust models to prepare for and prevent loss to follow-up by directing efforts toward individuals at higher risk. [ABSTRACT FROM AUTHOR]

Due to the wealth of exposome data from longitudinal cohort studies that is currently available, the need for methods to adequately analyze these data is growing. We propose an approach in which machine learning is used to identify longitudinal exposome-related predictors of health, and illustrate its potential through an application. Our application involves studying the relation between exposome and self-perceived health based on the 30-year running Doetinchem Cohort Study. Random Forest (RF) was used to identify the strongest predictors due to its favorable prediction performance in prior research. The relation between predictors and outcome was visualized with partial dependence and accumulated local effects plots. To facilitate interpretation, exposures were summarized by expressing them as the average exposure and average trend over time. The RF model's ability to discriminate poor from good self-perceived health was acceptable (Area-Under-the-Curve = 0.707). Nine exposures from different exposome-related domains were largely responsible for the model's performance, while 87 exposures seemed to contribute little to the performance. Our approach demonstrates that ML can be interpreted more than widely believed, and can be applied to identify important longitudinal predictors of health over the life course in studies with repeated measures of exposure. The approach is context-independent and broadly applicable. [ABSTRACT FROM AUTHOR]

This study investigated how changes in weather factors affect the prevalence of conjunctivitis using public big data in South Korea. A total of 1,428 public big data entries from January 2013 to December 2019 were collected. Disease data and basic climate/air pollutant concentration records were collected from nationally provided big data. Meteorological factors affecting eye diseases were identified using multiple linear regression and machine learning analysis methods such as extreme gradient boosting (XGBoost), decision tree, and random forest. The prediction model with the best performance was XGBoost (1.180), followed by multiple regression (1.195), random forest (1.206), and decision tree (1.544) when using root mean square error (RMSE) values. With the XGBoost model, province was the most important variable (0.352), followed by month (0.289) and carbon monoxide exposure (0.133). Other air pollutants including sulfur dioxide, PM10, nitrogen dioxides, and ozone showed low associations with conjunctivitis. We identified factors associated with conjunctivitis using traditional multiple regression analysis and machine learning techniques. Regional factors were important for the prevalence of conjunctivitis as well as the atmosphere and air quality factors. [ABSTRACT FROM AUTHOR]

Due to concealed initial symptoms, many diabetic patients are not diagnosed in time, which delays treatment. Machine learning methods have been applied to increase the diagnosis rate, but most of them are black boxes lacking interpretability. Rule extraction is usually used to turn on the black box. As the number of diabetic patients is far less than that of healthy people, the rules obtained by the existing rule extraction methods tend to identify healthy people rather than diabetic patients. To address the problem, a method for extracting reduced rules based on biased random forest and fuzzy support vector machine is proposed. Biased random forest uses the k-nearest neighbor (k-NN) algorithm to identify critical samples and generates more trees that tend to diagnose diabetes based on critical samples to improve the tendency of the generated rules for diabetic patients. In addition, the conditions and rules are reduced based on the error rate and coverage rate to enhance interpretability. Experiments on the Diabetes Medical Examination Data collected by Beijing Hospital (DMED-BH) dataset demonstrate that the proposed approach has outstanding results (MCC = 0.8802) when the rules are similar in number. Moreover, experiments on the Pima Indian Diabetes (PID) and China Health and Nutrition Survey (CHNS) datasets prove the generalization of the proposed method. [ABSTRACT FROM AUTHOR]

Solar energy-based technologies have developed rapidly in recent years, however, the inability to appropriately estimate solar energy resources is still a major drawback for these technologies. In this study, eight different artificial intelligence (AI) models namely; convolutional neural network (CNN), artificial neural network (ANN), long short-term memory recurrent model (LSTM), eXtreme gradient boost algorithm (XG Boost), multiple linear regression (MLR), polynomial regression (PLR), decision tree regression (DTR), and random forest regression (RFR) are designed and compared for solar irradiance prediction. Additionally, two hybrid deep neural network models (ANN-CNN and CNN-LSTM-ANN) are developed in this study for the same task. This study is novel as each of the AI models developed was used to estimate solar irradiance considering different timesteps (hourly, every minute, and daily average). Also, different solar irradiance datasets (from six countries in Africa) measured with various instruments were used to train/test the AI models. With the aim to check if there is a universal AI model for solar irradiance estimation in developing countries, the results of this study show that various AI models are suitable for different solar irradiance estimation tasks. However, XG boost has a consistently high performance for all the case studies and is the best model for 10 of the 13 case studies considered in this paper. The result of this study also shows that the prediction of hourly solar irradiance is more accurate for the models when compared to daily average and minutes timestep. The specific performance of each model for all the case studies is explicated in the paper. [ABSTRACT FROM AUTHOR]