Search

• Auto-encoder-enhanced Random Forest models with network biomarkers accurately predict IgAN prognosis. • Personalized IgAN treatment guidance system developed using network biomarkers. • System's performance on par with nephrologists, improving healthcare decision-making. The ongoing debate surrounding the use of immunosuppressive treatments for IgA nephropathy (IgAN) underscores the demand for personalized and effective strategies. Analyzed data from 807 IgAN patients over 5+ years using three methods: Random Forest with molecular biomarkers, network biomarkers with graph engineering, and an auto-encoder model. All models were trained using identical demographic, clinical, and pathological data, employing an 80–20 split for training and testing purposes. In the comprehensive assessment of IgAN prognosis, the Random Forest model, employing molecular biomarkers, demonstrated strong performance metrics (AUC = 0.83, sensitivity = 0.51, specificity = 0.96). However, traditional graph feature engineering on patient-specific networks outperformed these results with an AUC of 0.90, sensitivity of 0.64, and specificity of 0.94. The Auto-encoder model showed the best accuracy (AUC = 0.91, sensitivity = 0.46, specificity = 0.96). The findings highlighted the superior predictive capabilities of network biomarkers over molecular biomarkers for adverse renal outcome prediction in IgAN. Consequently, we integrated Auto-encoder-derived Network Biomarkers with Random Forest Models to enhance prognostic precision in diverse IgAN treatment scenarios. The prediction for the prognosis of patients receiving supportive care, glucocorticoid therapy, and immunosuppressant treatment yielded AUC values of 0.95, 0.96, and 1, respectively, indicating high specificity. Drawing from these insights, we pioneered the development of an innovative decision support model for IgAN treatment. This model demonstrated the ability to make medical decisions comparable to those by experienced nephrologists, enabling the customization of personalized disease management strategies. Our system accurately predicted IgAN prognosis and evaluated various treatment efficacies, aiding physicians in devising optimal therapeutic strategies for patients. [ABSTRACT FROM AUTHOR]

The dynamic and hierarchical nature of the functional brain network. The neural dynamical systems tend to converge to multiple attractors (stable fixed points or dynamical states) in long run. Little is known about how the changes in this brain dynamic "long-term" behavior of the connectivity flow of brain network in generalized anxiety disorder (GAD). This study recruited 92 patients with GAD and 77 healthy controls (HC). We applied a reachable probability approach combining a Non-homogeneous Markov model with transition probability to quantify all possible connectivity flows and the hierarchical structure of brain functional systems at the dynamic level and the stationary probability vector (10-step transition probabilities) to describe the steady state of the system in the long run. A random forest algorithm was conducted to predict the severity of anxiety. The dynamic functional patterns in distributed brain networks had larger possibility to converge in bilateral thalamus, posterior cingulate cortex (PCC), right superior occipital gyrus (SOG) and smaller possibility to converge in bilateral superior temporal gyrus (STG) and right parahippocampal gyrus (PHG) in patients with GAD compared to HC. The abnormal transition probability pattern could predict anxiety severity in patients with GAD. Small samples and subjects taking medications may have influenced our results. Future studies are expected to rule out the potential confounding effects. Our results have revealed abnormal dynamic neural communication and integration in emotion regulation in patients with GAD, which give new insights to understand the dynamics of brain function of patients with GAD. • We applied a Non-homogeneous Markov model to captured the spatio-temporal dynamic patterns underlying brain network. • Patients with GAD displayed abnormal transition probability during the steady state of the human brain functional network. • The abnormal dynamic pattern could well predict anxiety severity in patients with GAD. [ABSTRACT FROM AUTHOR]

Hydrogen, as a clean and versatile energy carrier, plays a pivotal role in addressing global energy challenges and transitioning towards sustainable energy systems. This study explores the convergence of machine learning (ML) for photocatalytic hydrogen evolution from sucrose solution using perovskite-type catalysts, namely LaFeO 3 (LFO) and graphene-supported LaFeO 3 (GLFO). This study pioneers the practical application of ML techniques, including Random Forests, LightGBM, and Bagging Regressor, to predict hydrogen yields in the presence of these photocatalysts. LFO and GLFO underwent a thorough characterization study to validate their successful preparation. Noteworthy, the highest hydrogen yield from the sucrose model solution was achieved using GLFO as 3.52 mmol/g cat. The optimum reaction conditions were experimentally found to be pH = 5.25, 0.15 g/L of catalyst amount, and 7.5 mM of HPC (hydrogen peroxide concentration). A pivotal contribution of this research lies in the practical application of ML models, culminating in the development of an ensemble model. This collaborative approach not only achieved an overall R2 of 0.92 but also demonstrated exceptional precision, as reflected in remarkably low error metrics. The mean squared logarithmic error (MSLE) was 0.0032, and the mean absolute error (MAE) was 0.049, underscoring the effectiveness of integrating diverse ML algorithms. This study advances both the understanding of photocatalytic hydrogen evolution and the practical implementation of ML in predicting intricate chemical reactions. [Display omitted] • Hydrogen, a clean energy carrier, tackles global challenges. • ML can predict hydrogen amounts evolved from sucrose solution. • GLFO achieves highest hydrogen yield at 3.52 mmol/gcat. • Ensemble of RF, LightGBM, and Bagging Regressor yields high accuracy with minimal errors. [ABSTRACT FROM AUTHOR]

Machine learning classifiers have been widely used for crop mapping; however, employing numerous features can reduce classifier accuracy. Feature selection is a widely recognized technique that addresses the issue of high dimensionality by choosing the most relevant subset of features with the highest degree of relevance and practicality. Therefore, the optimal combination of feature selection methods and classifiers for high-precision crop mapping remains an open problem. In this study, we introduce a novel hybrid feature selection approach to obtain the optimal subset of features for effective crop mapping. Initially, spectro-temporal remote sensing features from the dataset were ranked using two distinct feature selection techniques: Mutual Information (MI) and ReliefF filter. The most relevant features from each filter-based approach were combined into a union subset. Subsequently, the Gray Wolf Optimization (GWO), a metaheuristic optimization technique, was applied to enhance the feature set generated in the initial step. In the final stage, a random forest classifier leveraged the optimized feature subset for accurate crop type prediction. The effectiveness of the proposed approach was evaluated in Behiera governorate, Egypt. Comparative analysis against existing crop mapping methods demonstrated the superior performance of the proposed approach, achieving an accuracy rate of 82%. Furthermore, this approach showcased a significant reduction in the feature count, streamlining the set from 308 down to only 26 features. [ABSTRACT FROM AUTHOR]

• Proposed a Hybrid Model for Paddy Crop prediction based on phenological parameters. • CNN-LSTM attained RMSE value of 0.219 t/ha compared to other existing methods. • Random Forest method achieved highest accuracy of 96.6% in paddy classification. Rice is one of the predominant food sources to fulfill the dietary requirements of well-being in India. Therefore, accurate and timely paddy crop yield prediction is crucial to ensure the food security of the country. In this direction, the present study proposed a hybrid deep-learning method based on Conv-1D and LSTM layers using the classification-derived phenological with meteorological parameters for paddy crop yield prediction. The paddy crop classification has been conducted using high-resolution (10 m) multispectral imagery based on GPS coordinates collected during the paddy field visits to extract the phenological parameters for input to the prediction model. In this context, the efficiency of Random Forest, Naïve Bayes, SVM, and Gradient Tree boost classifiers was assessed. Furthermore, we have also analyzed the accuracy of Landsat-8, Sentinel-1 GRD, and Sentinel-2 satellite imagery in paddy crop classification based on area estimation. The Statistical Abstract of Haryana was utilized to validate the paddy crop area estimation and yield prediction. The classification outcomes showed that the Random Forest method attained the highest accuracy of 96.6 % compared to other GEE-based classifiers. The proposed Hybrid Deep learning approach achieved an RMSE value of 0.219 t/ha compared to CNN, LSTM, CNN-Bi-LSTM, and Regression techniques for crop yield prediction. The study conclusion highlighted that the sentinel-2 satellite imagery performed well and found that the proposed hybrid approach provided an alternative for paddy crop yield prediction. [ABSTRACT FROM AUTHOR]

• Synergistic use of drone, satellite, and machine learning technologies improves LAI estimation. • Robust validation and advanced performance metrics confirm model accuracy and precision. • Utilizing Vegetation Indices to enhance the precision of LAI quantification. In this study, we conducted a comprehensive assessment of Leaf Area Index (LAI) estimation using three distinct sources of satellite data: Sentinel-2 imagery, drone imagery (UAVs), and Mohammed VI satellite data. The main objective was to identify the most reliable and precise dataset for predicting LAI, with a focus on evaluating the performance of Random Forest models. For Sentinel-2 imagery, our Random Forest model achieves a robust R-squared (R2) value of 0.89, signifying a strong alignment between predicted and measured LAI values. The associated root-mean-square error (RMSE) is 0.4, indicating high predictive accuracy. In the context of UAVs, our Random Forest model excels, exhibiting an impressive R2 value of 0.93, highlighting a substantial correlation between predicted and measured LAI. The RMSE for drone imagery stands at 0.37, showcasing exceptional predictive accuracy. Finally, the Random Forest model trained on Mohammed VI satellite data yields an R2 value of 0.92, underlining its strong fit with measured LAI values. The RMSE for Mohammed VI imagery is 0.39, further underscoring the model's exceptional predictive accuracy. This comparative analysis underscores the importance of selecting the most suitable satellite data source for LAI estimation in Argania spinosa. UAV imagery emerges as the most accurate choice, closely followed by Mohammed VI Satellite and Sentinel-2 imagery. These findings offer valuable insights for effective monitoring of Argania spinosa and advancing sustainable land management practices in rural ecosystems. [ABSTRACT FROM AUTHOR]

Cognitive deficits are a core symptom of schizophrenia, but research on their neural underpinnings has been challenged by the heterogeneity in deficits' severity among patients. Here, we address this issue by combining logistic regression and random forest to classify two neuropsychological profiles of patients with high (HighCog) and low (LowCog) cognitive performance in two independent samples. We based our analysis on the cortical features grey matter volume (VOL), cortical thickness (CT), and mean curvature (MC) of N = 57 patients (discovery sample) and validated the classification in an independent sample (N = 52). We investigated which cortical feature would yield the best classification results and expected that the 10 most important features would include frontal and temporal brain regions. The model based on MC had the best performance with area under the curve (AUC) values of 76% and 73%, and identified fronto-temporal and occipital brain regions as the most important features for the classification. Moreover, subsequent comparison analyses could reveal significant differences in MC of single brain regions between the two cognitive profiles. The present study suggests MC as a promising neuroanatomical parameter for characterizing schizophrenia cognitive subtypes. [ABSTRACT FROM AUTHOR]

• Our random forest quantitative EEG (qEEG) classifier could classify delirium versus no delirium with an AUC of 0.76. • The most important delirium classification features were: theta peak frequency, relative alpha power and theta autocorrelation. • There was no evidence that qEEG features differed between postoperative and non-postoperative delirium. This exploratory study examined quantitative electroencephalography (qEEG) changes in delirium and the use of qEEG features to distinguish postoperative from non-postoperative delirium. This project was part of the DeltaStudy, a cross-sectional, multicenter study in Intensive Care Units (ICUs) and non-ICU wards. Single-channel (Fp2-Pz) four-minutes resting-state EEG was analyzed in 456 patients. After calculating 98 qEEG features per epoch, random forest (RF) classification was used to analyze qEEG changes in delirium and to test whether postoperative and non-postoperative delirium could be distinguished. An area under the receiver operating characteristic curve (AUC) of 0.76 (95% Confidence Interval (CI) 0.71–0.80) was found when classifying delirium with a sensitivity of 0.77 and a specificity of 0.63 at the optimal operating point. The classification of postoperative versus non-postoperative delirium resulted in an AUC of 0.50 (95%CI 0.38–0.61). RF classification was able to discriminate delirium from no delirium with reasonable accuracy, while also identifying new delirium qEEG markers like autocorrelation and theta peak frequency. RF classification could not distinguish postoperative from non-postoperative delirium. Single-channel EEG differentiates between delirium and no delirium with reasonable accuracy. We found no distinct EEG profile for postoperative delirium, which may suggest that delirium is one entity, whether it develops postoperatively or not. [ABSTRACT FROM AUTHOR]

Predictive models of psychotic symptoms could improve ecological momentary interventions by dynamically providing help when it is needed. Wearable sensors measuring autonomic arousal constitute a feasible base for predictive models since they passively collect physiological data linked to the onset of psychotic experiences. To explore this potential, we investigated whether changes in autonomic arousal predict the onset of hallucination spectrum experiences (HSE) and paranoia in individuals with an increased likelihood of experiencing psychotic symptoms. For 24 h of ambulatory assessment, 62 participants wore electrodermal activity and heart rate sensors and were provided with an Android smartphone to answer questions about their HSE-, and paranoia-levels every 20 min. We calculated random forests to detect the onset of HSEs and paranoia. The generalizability of our models was tested using leave-one-assessment-out and leave-one-person-out cross-validation. Leave-one-assessment-out models that relied on physiological data and participant ID yielded balanced accuracy scores of 80 % for HSE and 66 % for paranoia. Adding baseline information about lifetime experiences of psychotic symptoms increased balanced accuracy to 82 % (HSE) and 70 % (paranoia). Leave-one-person-out models yielded lower balanced accuracy scores (51 % to 58 %). Using passively collectible variables to predict the onset of psychotic experiences is possible and prediction models improve with additional information about lifetime experiences of psychotic symptoms. Generalizing to new individuals showed poor performance, so including personal data from a recipient may be necessary for symptom prediction. Completely individualized prediction models built solely with the data of the person to be predicted might increase accuracy further. [ABSTRACT FROM AUTHOR]

To predict renal tumour growth patterns in von Hippel–Lindau syndrome by utilising radiomic features to assist in developing personalised surveillance plans leading to better patient outcomes. The study evaluated 78 renal tumours in 55 patients with histopathologically-confirmed clear cell renal cell carcinomas (ccRCCs), which were segmented and radiomics were extracted. Volumetric doubling time (VDT) classified the tumours into fast-growing (VDT <365 days) or slow-growing (VDT ≥365 days). Volumetric and diametric growth analyses were compared between the groups. Multiple logistic regression and random forest classifiers were used to select the best features and models based on their correlation and predictability of VDT. Fifty-five patients (mean age 42.2 ± 12.2 years, 27 men) with a mean time difference of 3.8 ± 2 years between the baseline and preoperative scans were studied. Twenty-five tumours were fast-growing (low VDT, i.e., <365 days), and 53 tumours were slow-growing (high VDT, i.e., ≥365 days). The median volumetric and diametric growth rates were 1.71 cm3/year and 0.31 cm/year. The best feature using univariate analysis was wavelet-HLL_glcm_ldmn (area under the receiver operating characteristic [ROC] curve [AUC] of 0.80, p< 0.0001), and with the random forest classifier, it was log-sigma-0-5-mm-3D_glszm_ZonePercentage (AUC: 79). The AUC of the ROC curves using multiple logistic regression was 0.74, and with the random forest classifier was 0.73. Radiomic features correlated with VDT and were able to predict the growth pattern of renal tumours in patients with VHL syndrome. • Median renal tumor doubling time in VHL syndrome was 538 days (278–539). • Median diametric and volumetric growth rates were 0.31 cm/year (0.17–0.51) and 1.71 cm3/year (0.8–4.2). • Radiomic features fairly predicted tumor growth patterns in VHL syndrome patients. • Models employed for feature selection: multiple logistic regression and random forest. • Potential of radiomics to improve outcomes by personalized surveillance strategies. [ABSTRACT FROM AUTHOR]

[Display omitted] • CC16 is an anti-inflammatory, immunomodulatory protein expressed in respiratory club cells. • Severe abdominal obesity and arterial hypertension robustly decrease serum CC16. • ACEi/ARBs, uricosurics and chronic heart failure robustly increase serum CC16. • Effects might be mediated by adipose tissue inflammation as well as RAAS and uric acid disturbance. • Findings indicate a complex interplay of the metabolic, respiratory and cardiovascular system. Clara cell 16-kDa protein (CC16) is an anti-inflammatory, immunomodulatory secreted pulmonary protein with reduced serum concentrations in obesity according to recent data. Studies focused solely on bodyweight, which does not properly reflect obesity-associated implications of the metabolic and reno-cardio-vascular system. The purpose of this study was therefore to examine CC16 in a broad physiological context considering cardio-metabolic comorbidities of primary pulmonary diseases. CC16 was quantified in serum samples in a subset of the FoCus (N = 497) and two weight loss intervention cohorts (N = 99) using ELISA. Correlation and general linear regression analyses were applied to assess CC16 effects of lifestyle, gut microbiota, disease occurrence and treatment strategies. Importance and intercorrelation of determinants were validated using random forest algorithms. CC16 A38G gene mutation, smoking and low microbial diversity significantly decreased CC16. Pre-menopausal female displayed lower CC16 compared to post-menopausal female and male participants. Biological age and uricosuric medications increased CC16 (all p < 0.01). Adjusted linear regression revealed CC16 lowering effects of high waist-to-hip ratio (est. −11.19 [−19.4; −2.97], p = 7.99 × 10−3), severe obesity (est. −2.58 [−4.33; −0.82], p = 4.14 × 10−3) and hypertension (est. −4.31 [−7.5; −1.12], p = 8.48 × 10−3). ACEi/ARB medication (p = 2.5 × 10−2) and chronic heart failure (est. 4.69 [1.37; 8.02], p = 5.91 × 10−3) presented increasing effects on CC16. Mild associations of CC16 were observed with blood pressure, HOMA-IR and NT-proBNP, but not manifest hyperlipidemia, type 2 diabetes, diet quality and dietary weight loss intervention. A role of metabolic and cardiovascular abnormalities in the regulation of CC16 and its modifiability by behavioral and pharmacological interventions is indicated. Alterations by ACEi/ARB and uricosurics could point towards regulatory axes comprising the renin-angiotensin-aldosterone system and purine metabolism. Findings altogether strengthen the importance of interactions among metabolism, heart and lungs. [ABSTRACT FROM AUTHOR]

Adolescent non-suicidal self-injury (NSSI) is a major public health issue. Family factors are significantly associated with NSSI in adolescents, while studies on forecasting NSSI at the family level are still limited. In addition to regression methods, machine learning (ML) techniques have been recommended to improve the accuracy of family-level risk prediction for NSSI. Using a dataset of 7967 students and their primary caregivers from a cross-sectional study, logistic regression model and random forest model were used to test the forecasting accuracy of NSSI predictions at the family level. Cross-validation was used to assess model prediction performance, including the area under the receiver operator curve (AUC), precision, Brier score, accuracy, sensitivity, specificity, positive predictive value and negative predictive value. The top three important family-related predictors within the random forest algorithm included family function (importance:42.66), family conflict (importance:42.18), and parental depression (importance:27.21). The most significant family-related risk predictors and protective predictors identified by the logistic regression model were family history of mental illness (OR:2.25) and help-seeking behaviors of mental distress from parents (OR:0.65), respectively. The AUCs of the two models, logistic regression and random forest, were 0.852 and 0.835, respectively. The key limitation is that this cross-sectional survey only enabled the authors to examine predictors that were considered to be proximal rather than distal. These findings highlight the significance of family-related factors in forecasting NSSI in adolescents. Combining both conventional statistical methods and ML methods to improve risk assessment of NSSI at the family level deserves attention. • The most important familial predictor within the random forest algorithm was family function. • The logistic regression model found family history of mental illness as the foremost familial risk predictor for NSSI. • Help-seeking behaviors of mental distress from parents or others in the family were protective predictors of NSSI. • Maternal education less than high school was a protective predictor of NSSI. • Both models resulted in an overall positive ability to predict family-related factors of NSSI. [ABSTRACT FROM AUTHOR]

The estimation of material performance plays a crucial role in practical life, enabling the rational allocation of time and resources while enhancing the practical application of materials. Therefore, this study investigates the analytical and predictive capabilities of five machine learning (ML) algorithms (the Random Forest algorithm – RF, the Adaboost algorithm – AB, the Gradient Boosting algorithm – GB, the Extra Trees algorithm – ET and linear models – LM) for comprehensive performance parameters of wood ceramics (ammonia nitrogen adsorption quantity - Q , open porosity - P , burn-off , and hardness). In the analysis of model prediction capabilities, five key statistical parameters (Root Mean Square Error - RMSE , Mean Square Error - MSE , Coefficient of Correlation - R , Determination Coefficient - R 2 , and Mean Absolute Relative Error - MARE) were calculated. The results indicate the following: (1) Among various ML models, the GB model exhibits the most superior performance with an R 2 ≥ 0.9594. (2) However, in predicting the four performance parameters of wood ceramics, the AB model with an R 2 ranging from 0.0078 to 0.45, indicating notably poor predictive capabilities. Despite integrating the Support Vector Regression (SVR) module, no enhancement in predictive accuracy was observed. (3) While forecasting the four performance parameters of wood ceramics, the LM model demonstrates predictive accuracy relatively akin to that of the RF and ET models among the ML models. (4) Though the feature importance scores vary across distinct input variable models, there is a consistent trend of change. (5) Variables B and D exhibit some level of correlation with other variables. In summary, the results of this study suggest that, for the analysis and prediction of wood ceramic performance, the GB model among the five regression models demonstrates outstanding simulation performance. [ABSTRACT FROM AUTHOR]

In the neurodevelopmental model of schizophrenia, minor physical anomalies (MPAs) are considered neurodevelopmental markers of schizophrenia. To date, there has been no research to evaluate the interaction between MPAs. Our study built and used a machine learning model to predict the risk of schizophrenia based on measurements of MPA items and to investigate the potential primary and interaction effects of MPAs. The study included 470 patients with schizophrenia and 354 healthy controls. The models used are classical statistical model, Logistic Regression (LR), and machine leaning models, Decision Tree (DT) and Random Forest (RF). We also plotted two-dimensional scatter diagrams and three-dimensional linear/quadratic discriminant analysis (LDA/QDA) graphs for comparison with the DT dendritic structure. We found that RF had the highest predictive power for schizophrenia (Full-training AUC = 0.97 and 5-fold cross-validation AUC = 0.75). We identified several primary MPAs, such as the mouth region, high palate, furrowed tongue, skull height and mouth width. Quantitative MPA analysis indicated that the higher skull height and the narrower mouth width, the higher the risk of schizophrenia. In the interaction, we further identified that skull height and mouth width, furrowed tongue and skull height, high palate and skull height, and high palate and furrowed tongue, showed significant two-item interactions with schizophrenia. A weak three-item interaction was found between high palate, skull height, and mouth width. In conclusion, we found that the two machine learning methods showed good predictive ability in assessing the risk of schizophrenia using the primary and interaction effects of MPAs. [ABSTRACT FROM AUTHOR]

Urban trees are not only a core component of natural infrastructure but also an effective way to mitigate urban heat with nature-based solutions. Comprehensively revealing the cooling effects of trees and their drivers is valuable for enhancing urban climate resilience and promoting sustainable development. While existing studies have investigated the cooling effects of two-dimensional characteristics of trees, there has been limited consideration of the effects of vertical structure, especially in various climatic zones across the globe. In this study, we employed the Google Earth Engine cloud platform and the random forest algorithm to comprehensively assess the impact of three-dimensional (3D) characteristics of trees on land surface temperature across 596 cities worldwide. Results suggest that LST is generally lower in tree-covered areas than their surrounding built-up land, especially in the summer, with an average decrease of about 2.13 °C. We also found a significant negative correlation between tree canopy height and LST (∼−0.83). Specifically, the mean LST decreases by about 0.16 ℃ for every 1m increase in tree height. Globally, the average cooling intensity of trees is 1.86 °C, and is about 1.06 °C higher in summer than in winter. It is worth noting that during winter, the cooling effect of trees is more pronounced closer to the equator. In addition, the 3D characteristics of trees contribute more significantly to cooling intensity compared to their surrounding environmental factors. This study not only fills a global knowledge gap regarding the impact of 3D features of trees on LST, but also provides valuable insights into urban planning and management in response to climate change. [ABSTRACT FROM AUTHOR]

Ionospheric delay is of concern for trans -ionospheric radio communication, especially for the navigation systems relying on these satellite signals. Most of the ionospheric delays are estimated to a degree of first-order using dual frequency global navigation satellite system (GNSS) receivers except during irregularities and equatorial plasma bubbles. The plasma bubbles are observed as a decrease or reduction in total electron content (TEC) as a result of large-scale irregularities that are generated at the equatorial ionosphere. These plasma bubbles can be detected from TEC values visually or by using mathematical algorithms. The mathematical algorithms may have limitations based on assumptions made for the current dataset. Therefore, various machine learning (ML) techniques were tried by training them with selected TEC depletions that are verified for accuracy. From this study, the Random Forest Method (RFM) has performed well compared to other ML methods. The RFM is trained to use for the detection of TEC depletions from the Indian low-latitude region. The training accuracy obtained is 97.6%, with a minimum classification error of 0.023%. The result obtained from the confusion matrix ascertained that the proportion of positively classified cases that are truly positive that is the positive predictive value (PPV) is 96.8%. These statistical results are validated after plotting the observed TEC depletion patch obtained from the ML method. There are cases in which depletions detected by the ML method are appreciable over the mathematical algorithm. The ML technique once trained will not have inherent limitations, as there are no assumptions or threshold values needed to set as required by most of the mathematical algorithms. Thus, the results are encouraging and have scope for further improvement and advancement. [ABSTRACT FROM AUTHOR]

Cohesion between team members is critical for surgical performance. Our previous study has shown that the experience of working together (measured by Team Familiarity Score, TFS) helps reduce procedure time (PT). However, that conclusion was found in a relatively small sample size. With a large dataset including mixed general surgical procedures, we hypothesize that team familiarity makes a significant contribution to the improvement of team performance in complex cases, rather than in medium or basic surgical cases, measured by the procedure time, length of hospital stays (LOS), and surgical cost (COST). Patient demographics, operation, and patient outcome data of 922 general surgery cases were included. The cases were divided into three subgroups, including basic, medium, and complex surgical procedures. TFS and an Index of Difficulty of Surgery (IDS) were calculated for each procedure. Simple linear regression and random forest regressions were performed to analyze the association between surgical outcomes and all included independent variables (TFS, IDS, patient age, patient weight, and team size). When applied to complex cases, procedure time (r = −0.21) and cost (r = −0.23) dropped as TFS increases. In basic and medium surgical cases, increasing team familiarity failed to shorten the procedure time on average. Team familiarity is more important in complex cases because there is greater potential for improvement through team collaboration compared to basic and medium cases. Caution will be needed when applying team familiarity scores for examining surgical team performance in large databases with skewed to basic surgical cases. • We calculated familiarity scores among surgical team members. • Higher team familiarity reduces procedure time and cost in complex cases. • Team familiarity has minor impact in Basic and Medium surgeries. • Precaution is needed when using familiarity scores in surgical teams. [ABSTRACT FROM AUTHOR]

Extracting knowledge from data streams requires fast incremental algorithms that are able to handle unlimited processing and ever-changing data with finite memory. A strategy for this challenge is the use of ensembles owing to their ability to tackle concept drift and achieve highly accurate predictions. However, ensembles often require a lot of computational resources. In this study, we propose a novel ensemble-based classification algorithm for data streams, Adaptive Regularized Ensemble (ARE), with low demand for computational resources. The algorithm combines strategies that contribute to high prediction accuracy using only incorrectly classified instances into the training step, random-sized feature subspace for each ensemble element and classifier selection for final ensemble voting. After an extensive experimental study, ARE exhibited high predictive performance and outperformed state-of-the-art ensembles on data streams for real and synthetic datasets while requiring a low processing time and memory consumption. • Classification algorithm for data streams fast and with low memory consumption. • Ensemble that incorporates only instances incorrectly classified in the training step. • Ensemble with random feature subspace size. • Classifier selection for final ensemble voting. [ABSTRACT FROM AUTHOR]

Methane is the second largest anthropogenic greenhouse gas, and changes in atmospheric methane concentrations can reflect the dynamic balance between its emissions and sinks. Therefore, the monitoring of CH 4 concentration changes and the assessment of underlying driving factors can provide scientific basis for the government's policy making and evaluation. China is the world's largest emitter of anthropogenic methane. However, due to the lack of ground-based observation sites, little work has been done on the spatial-temporal variations for the past decades and influencing factors in China, especially for areas with high anthropogenic emissions as Central and Eastern China. Here to quantify atmospheric CH 4 enhancements trends and its driving factors in Central and Eastern China, we combined the most up-to-date TROPOMI satellite-based column CH 4 (xCH 4) concentration from 2018 to 2022, anthropogenic and natural emissions, and a random forest-based machine learning approach, to simulate atmospheric xCH 4 enhancements from 2001 to 2018. The results showed that (1) the random forest model was able to accurately establish the relationship between emission sources and xCH 4 enhancement with a correlation coefficient (R²) of 0.89 and a root mean-square error (RMSE) of 11.98 ppb; (2)The xCH 4 enhancement only increased from 48.21±2.02 ppb to 49.79±1.87 ppb from the year of 2001 to 2018, with a relative change of 3.27%±0.13%; (3) The simulation results showed that the energy activities and waste treatment were the main contributors to the increase in xCH 4 enhancement, contributing 68.00% and 31.21%, respectively, and the decrease of animal ruminants contributed -6.70% of its enhancement trend. [Display omitted] [ABSTRACT FROM AUTHOR]

The impact of occupational stress and work environment fitness on mental health disparities between physicians and nurses are not well understood. This study aims to identify and rank key determinants of mental health in physicians and nurses in China and compare the differences in their impact on mental health between physicians and nurses. A large cross-sectional survey with multistage cluster sampling was conducted. The survey included the Self-Rating Anxiety Scale (SAS Scale), the Center for Epidemiologic Studies Depression Scale (CES-D Scale), the Maslach Burnout Inventory–General Survey (MBI-GS) and the Person-Environment (PE) Fit. We applied a principled, machine learning-based variable selection algorithm, using random forests, to identify and rank the determinants of the mental health in physicians and nurses. In our study, we analyzed a sample of 9964 healthcare workers, and 2729 (27 %) were physicians. The prevalence of anxiety and depressive disorders among physicians and nurses was 31.0 % and 53.3 %, 30.8 % and 47.9 %, respectively. Among physicians with anxiety disorder, we observed a higher likelihood of cynicism, emotional exhaustion, reduced personal accomplishment, and poor organization fitness, job fitness, group fitness, and supervisor fitness, in order of importance. When comparing the effects on depressive disorder in physicians, group fitness and supervisor fitness did not have significant impacts. For nurses, emotional exhaustion had a more significant effect on depressive disorder compared to cynicism. Supervisor fitness did not have a significant impact on anxiety disorder in nurses. Cross-sectional design, self-reporting screening scales. Compared to individual and hospital characteristics, the primary factors influencing mental health disorders are occupational burnout and the compatibility of the work environment. Additionally, the key determinants of depressive and anxiety disorders among doctors and nurses exhibit slight variations. Employing machine learning methods proves beneficial for identifying determinants of mental health disorders among physicians and nurses in China. These findings could help improve policymaking aimed at addressing the mental well-being of healthcare professionals. • Study uses rigorous methodology to identify mental health determinants in physicians and nurses. Informs debate on improving overall mental health and addressing variation in healthcare. • Top determinants of physician anxiety: cynicism, emotional exhaustion, reduced personal accomplishment, and poor organization fitness, job fitness, group fitness, and supervisor fitness. • Unlike anxiety, group and supervisor fitness not significantly linked to depressive disorder in physicians. • Top determinants of nurse anxiety: emotional exhaustion, cynicism, reduced personal accomplishment, and poor job fitness, organization fitness, and group fitness. • Supervisor fitness emerged as a key determinant significantly associated with depressive symptoms in nurses; organization fitness crucial, identified as the most impactful work environment factor for mental health outcomes. [ABSTRACT FROM AUTHOR]

As primary Sjogren's syndrome (pSS) primarily affects the salivary glands, saliva can serve as an indicator of the glands' pathophysiology and the disease's status. This study aims to illustrate the salivary proteomic profiles of pSS patients and identify potential candidate biomarkers for diagnosis. The discovery set contained 49 samples (24 from pSS and 25 from age- and gender-matched healthy controls [HCs]) and the validation set included 25 samples (12 from pSS and 13 from HCs). Totally 36 pSS patients and 38 HCs were centrally randomized into the discovery set or to the validation set at a 2:1 ratio. Unstimulated whole saliva samples from pSS patients and HCs were analyzed using a data-independent acquisition (DIA) strategy on a 2D LC–HRMS/MS platform to reveal differential proteins. The crucial proteins were verified using DIA analysis and annotated using gene ontology (GO) and International Pharmaceutical Abstracts (IPA) analysis. A prediction model for SS was established using random forests. A total of 1,963 proteins were discovered, and 136 proteins exhibited differential representation in pSS patients. The bioinformatic research indicated that these proteins were primarily linked to immunological functions, metabolism, and inflammation. A panel of 19 protein biomarkers was identified by ranking order based on P -value and random forest algorichm, and was validated as the predictive biomarkers exhibiting good performance with area under the curve (AUC) of 0.817 for discovery set and 0.882 for validation set. The candidate protein panel discovered may aid in pSS diagnosis. Salivary proteomic analysis is a promising non-invasive method for prognostic evaluation and early and precise treatments for pSS patients. DIA offers the best time efficiency and data dependability and may be a suitable option for future research on the salivary proteome. [ABSTRACT FROM AUTHOR]

The escalating global threat of forest fires, driven by global warming, requires the development of effective prediction systems to mitigate damages. This research focuses on Madhya Pradesh (MP) and Chhattisgarh (CG) states in central India, where forest fire risk has become particularly pronounced. The primary objectives of the study are to quantify and map the spatial and temporal dynamics of forest fires over the period 2001 to 2020, and to predict future fire risks using satellite derived datasets and machine learning techniques. Through a long-term analysis, the study revealed an alarming increase in the number of forest fire incidents in MP and CG. From an average of 1200 and 1000 during 2001 to 2005, the incidents increased to 2800 and 2100 during 2016 to 2020, in MP and CG respectively. To predict forest fire risk, Random Forest machine learning algorithm was adopted utilizing various satellite derived climatic, topographical, and ecological parameters such as temperature, precipitation, solar radiation, NDVI, soil moisture, litter availability, evapotranspiration and terrain parameters (at monthly scale for 20 years). While forecasting fire probability for 2018–2020, the model achieves high accuracy rate of 86.46 % in MP and 93.78 % in CG. The results highlight significant forest fire likelihood regions in the central MP and the Southern CG, identifying areas requiring enhanced fire management strategies. This study has revealed that NDVI and rainfall have played a positive role in restricting the forest fire, and their negative anomaly amplified the fire risk. The study would help forest planners and administrators to characterise vulnerable areas and prioritise their conservation provisions. [ABSTRACT FROM AUTHOR]

Underground hydrogen storage (UHS) offers a promising approach for the storage of significant volumes of hydrogen gas (H 2) within deep geological formations, which can later be utilized for energy generation when necessary. Interfacial tension (IFT) between H 2 and the formation brine plays a vital role in influencing the distribution of H 2 at the pore scale and, ultimately, the storage capacity. In this research, we developed four intelligent models: Decision Trees (DT), Random Forests (RF), Support Vector Machines (SVM), and Multi-Layer Perceptron (MLP). These models were designed to predict the IFT utilizing pressure, temperature, and molality. Additionally, we fine-tuned three explicit correlations previously developed in our research. To assess the influence of each parameter on IFT, we conducted a comprehensive analysis of raw data to exclude doubtful samples. This was followed by rigorous model development, including hyperparameter tuning, and finally, an examination of developed models using testing data. The results clearly demonstrate the superiority of the RF model, achieving high accuracy and reliability with coefficients of determination (R2), root mean square error (RMSE), and average absolute relative deviation (AARD) values of 0.96, 1.50, and 1.84 %, respectively. The exemplary performance of the RF model can be attributed to its inherent characteristics. The ensemble approach of RF excels in capturing complex data relationships, thereby enhancing predictive accuracy and solidifying its superiority over other models in this study. Furthermore, the feature importance analysis revealed that temperature has the most significant influence, followed by molality and pressure. Moreover, we assessed the predictive accuracy of these models through external testing using data not used in the initial training and testing stages. Our study highlights the exceptional predictive power of the RF model, emphasizing the practical importance of selecting RF for enhanced accuracy and reliability. The proposed method shows significant potential for industrial applications, especially in optimizing underground H 2 storage. • Various ML techniques were employed to calculate the IFT in H 2 -brine systems. • Different statistical metrics were employed to assess the models' performance. • The developed ML approaches showed acceptable predictions of the IFT values. • The RF based model outperformed other intelligent models. • The reliability and accuracy of models were demonstrated using external testing. [ABSTRACT FROM AUTHOR]

Radial fans play a critical role as indispensable turbomachines in various industrial sectors. However, the conventional manufacturing process for these fans is often characterized by its resource-intensive and time-consuming nature. Traditionally, computational fluid dynamics (CFD) has been the go-to method for predicting and analyzing the performance of radial fans at different geometrical and operational conditions. Yet, in recent years, the rapid advancements in machine learning (ML) and deep learning (DL) techniques, particularly the rise of artificial neural networks (ANNs), have propelled significant progress in the field of predicting and optimizing the performance of radial fans. The present study aims to analyze the performance of a radial fan through a comprehensive experimental investigation and a meticulous three-dimensional numerical simulation. Subsequently, in order to predict the off-design performance of the fan, an extensive set of numerical simulations is conducted at various volumetric flow rates and rotational speeds. These simulations are used to analyze the fan performance and identify the most efficient operating condition. Moreover, the simulations serve as inputs for a finely-tuned ANN architecture. The predictive accuracy of the ANN model for both interpolation and extrapolation cases is then compared against two alternative techniques, namely support vector machine (SVM) and random forest (RF). The results explicitly highlight the superiority of the ANN model in terms of its predictive accuracy, thereby solidifying its position as the most reliable method for predicting the performance of radial fans. [ABSTRACT FROM AUTHOR]

• Adherence to 12 selected guidelines was higher at pediatric than adult care centers. • Adherence to more guidelines predicted lower P. aeruginosa incidence rates (Pa-IR). • Macrolide prescription predicted lower and glucose tolerance test higher Pa-IR. Care guidelines for cystic fibrosis (CF) have been developed to enhance consistent care and to improve health outcomes. We determined if adherence to CF care guidelines predicted P. aeruginosa incidence rates (Pa-IR) at U.S. CF centers in 2018. This cross-sectional CF Foundation Patient Registry study included 82 adult and 132 pediatric centers. Adherence to 12 guidelines was defined categorically (guideline met) or as a continuous measure (proportion of patients being treated/evaluated per guideline). Association of adherence to individual guidelines with Pa-IR, accounted for center and patient characteristics relevant to Pa-IR and were modeled using random forests and weighted-least-squares (WLS) analyses. The mean Pa-IR was 0.2 cases/patient-years at risk (SE 0.0074) for all centers combined. Guideline adherence was lowest for ≥4 bacterial cultures/year (54% of centers) and annual oral glucose tolerance test (OGTT) (48% of centers), and highest for annual non-tuberculous mycobacteria (NTM) sputum culture (98%). The mean number of guidelines met was 6.7 and higher for pediatric (7.3) than adult (5.6) centers, (p <0.001). The number of guidelines met correlated negatively with Pa-IR (β=-0.007, p = 0.043). Macrolide prescription and annual OGTT per guideline were associated with lower and higher Pa-IR, respectively. Centers with lower center-wide lung function, higher proportion of pwCF with low body-mass index, and location in the Southwest had higher Pa-IR. Overall adherence to guidelines was high except for performing ≥4 bacterial cultures/year and OGTT. Higher Pa-IR was associated with center characteristics and lower guideline adherence. The lower Pa-IR with greater adherence to guidelines suggests that focusing on quality care can positively impact Pa-IR. [ABSTRACT FROM AUTHOR]

Many predictive models for estimating clinical outcomes after spine surgery have been reported in the literature. However, implementation of predictive scores in practice is limited by the time-intensive nature of manually abstracting relevant predictors. In this study, we designed natural language processing (NLP) algorithms to automate data abstraction for the thoracolumbar injury classification score (TLICS). We retrieved the radiology reports of all Mayo Clinic patients with an International Classification of Diseases, 9th or 10th revision, code corresponding to a fracture of the thoracolumbar spine between January 2005 and October 2020. Annotated data were used to train an N-gram NLP model using machine learning methods, including random forest, stepwise linear discriminant analysis, k-nearest neighbors, and penalized logistic regression models. A total of 1085 spine radiology reports were included in our analysis. Our dataset included 483 compression, 401 burst, 103 translational/rotational, and 98 distraction fractures. A total of 103 reports had documented an injury of the posterior ligamentous complex. The overall accuracy of the random forest model for fracture morphology feature detection was 76.96% versus 65.90% in the stepwise linear discriminant analysis, 50.69% in the k-nearest neighbors, and 62.67% in the penalized logistic regression. The overall accuracy to detect posterior ligamentous complex integrity was highest in the random forest model at 83.41%. Our random forest model was implemented in the backend of a web application in which users can dictate reports and have TLICS features automatically extracted. We have developed a machine learning NLP model for extracting TLICS features from radiology reports, which we deployed in a web application that can be integrated into clinical practice. [ABSTRACT FROM AUTHOR]

Temporal lobe epilepsy is the most common reason behind drug-resistant seizures and temporal lobectomy (TL) is performed after all other efforts have been taken for a Temporal lobe epilepsy. Our study aims to develop multiple machine learning (ML) models capable of predicting postoperative outcomes following TL surgery. Data from the American College of Surgeons National Surgical Quality Improvement Program database identified patients who underwent TL surgery. We focused on 3 outcomes: prolonged length of stay (LOS), nonhome discharges, and 30-day readmissions. Six ML algorithms, TabPFN, XGBoost, LightGBM, Support Vector Machine, Random Forest, and Logistic Regression, coupled with the Optuna optimization library for hyperparameter tuning, were tested. Models with the highest area under the receiver operating characteristic (AUROC) values were included in the web application. SHapley Additive exPlanations was used to evaluate importance of predictor variables. Our analysis included 423 patients. Of these patients, 111 (26.2%) experienced prolonged LOS, 33 (7.8%) had nonhome discharges, and 29 (6.9%) encountered 30-day readmissions. The top-performing models for each outcome were those built with the Random Forest algorithm. The Random Forest models yielded AUROCs of 0.868, 0.804, and 0.742 in predicting prolonged LOS, nonhome discharges, and 30-day readmissions, respectively. Our study uses ML to forecast adverse postoperative outcomes following TL. We developed accessible predictive models that enhance prognosis prediction for TL surgery. Making ML models available for this purpose represents a significant advancement in shifting toward a more patient-centric, data-driven paradigm. [ABSTRACT FROM AUTHOR]

Data science approaches have increasingly been used in behavioral health research and may be useful for addressing social factors contributing to disparities in health status. This study evaluated the importance of cultural stress-related factors in classifying depression and post-traumatic stress disorder (PTSD) among adult survivors (N = 319) of Hurricane Maria who migrated from Puerto Rico to the United States mainland. We evaluated the performance of random forests (RF) and logistic regression (LR) for classifying PTSD and depression. Models included demographic, hurricane exposure, and migration-related cultural stress variables. We inspected area under the receiver operating characteristic curve (AUC), accuracy, balanced accuracy, F1 score, precision, recall, and specificity. Negative context of reception and language-related stressors were moderately important for accurately classifying depression and PTSD. For classifying depression, RF showed higher accuracy, balanced accuracy, specificity, precision, and F1. For classifying PTSD, RF showed higher accuracy, specificity, precision, and F1. A more thorough classification model would also include biomarkers (e.g., of allostatic load), family, community, or neighborhood-level attributes. Findings may not generalize to other groups who have experienced crisis-related migration. Findings underscore the importance of culturally and linguistically appropriate and trauma-informed clinical services for recent migrants. Use of assessments to identify pre-migration and post-migration stressors could inform clinical practice with migrants presenting with behavioral health-related difficulties. • Random Forest and Logistic Regression models performed acceptably in classifying depression and PTSD. • Random Forest models had lower AUC but higher precision, specificity, and F1 than logistic regression models. • Language-related stress and negative context of reception were useful for distinguishing migrants with probable mental health conditions. [ABSTRACT FROM AUTHOR]

Research has shown a link between depression risk and how gamers form relationships with their in-game figure of representation, called avatar. This is reinforced by literature supporting that a gamer's connection to their avatar may provide broader insight into their mental health. Therefore, it has been argued that if properly examined, the bond between a person and their avatar may reveal information about their current or potential struggles with depression offline. To examine whether the connection with an individuals' avatars may reveal their risk for depression, longitudinal data from 565 adults/adolescents (Mage = 29.3 years, SD = 10.6) were evaluated twice (six months apart). Participants completed the User-Avatar-Bond [UAB] scale and Depression Anxiety Stress Scale to measure avatar bond and depression risk. A series of tuned and untuned artificial intelligence [AI] classifiers analyzed their responses concurrently and prospectively. This allowed the examination of whether user-avatar bond can provide cross-sectional and predictive information about depression risk. Findings revealed that AI models can learn to accurately and automatically identify depression risk cases, based on gamers' reported UAB, age, and length of gaming involvement, both at present and six months later. In particular, random forests outperformed all other AIs, while avatar immersion was shown to be the strongest training predictor. Study outcomes demonstrate that UAB can be translated into accurate, concurrent, and future, depression risk predictions via trained AI classifiers. Assessment, prevention, and practice implications are discussed in the light of these results. [ABSTRACT FROM AUTHOR]

The sulfur content at sulfide saturation (SCSS) of a silicate melt can regulate the stability of sulfides and, therefore, chalcophile elements' behaviors in planetary magma oceans. Many studies have reported high-pressure experiments to determine SCSS using either linear or exponential regressions to parameterize the thermodynamics of the system. Although these empirical equations describe the effects of different parameters on SCSS, they perform poorly when predicting laboratory measurements. Here, we compiled 542 published analyses of experiments performed on a range of sulfide and silicate compositions at varying P - T conditions (<24 GPa, <2673 K). Using empirical equations, linear regression, Random Forest algorithms, and a hybrid algorithm employing empirical fits to P - T conditions and the Random Forest algorithm for compositions, we developed several SCSS models and compared them to laboratory measurements. The Random Forest and hybrid models (R 2 = 0.82–0.91, mean average error [MAE] < 746 ppmw S, residual mean standard error [RMSE] < 972 ppmw S), significantly outperform previous empirical models (R 2 = 0.28–0.69, MAE = 622–1,170 ppmw S, RMSE = 1,070–1,744 ppmw S), whereas linear regression performs moderately well, i.e., between the classic and machine learning models. We applied our hybrid model to predict SCSS during magma ocean solidification on Earth, Mars, and the Moon, and we compared our model results to expected S contents in the residual magma oceans calculated by mass balance. Our results confirm that during early accretion, sulfides precipitated from magma oceans and into the outer cores of Earth and Mars, but not the Moon. Subsequently, once the respective magma oceans began precipitating minerals with increasingly FeO-rich and SiO 2 -, Al 2 O 3 -, and MgO-depleted compositions, the increasing S concentration in the residual magma was offset by temperature and compositional effects on SCSS, preventing sulfide precipitation during intermediate stages of crystallization. Sulfides precipitated late during magma ocean crystallization, but failed to percolate through the underlying crystalline mantle, significantly contributing to the modern bulk-silicate sulfur abundances of Earth, Mars, and the Moon. Our calculations suggest that late-stage sulfide precipitation occurred at shallow depths of 120–220 km, 40–320 km, and < 10 km in the magma oceans of Earth, Mars, and the Moon, respectively. [ABSTRACT FROM AUTHOR]

Our purpose was to analyze the effect on gastrointestinal (GI) toxicity models when their dose-volume metrics predictors are derived from segmentations of the peritoneal cavity after different contouring approaches. A random forest machine learning approach was used to predict acute grade ≥3 GI toxicity from dose-volume metrics and clinicopathologic factors for 246 patients (toxicity incidence = 9.5%) treated with definitive chemoradiation for squamous cell carcinoma of the anus. Three types of random forest models were constructed based on different bowel bag segmentation approaches: (1) physician-delineated after Radiation Therapy Oncology Group (RTOG) guidelines, (2) autosegmented by a deep learning model (nnU-Net) following RTOG guidelines, and (3) autosegmented but spanning the entire bowel space. Each model type was evaluated using repeated cross-validation (100 iterations; 50%/50% training/test split). The performance of the models was assessed using area under the precision-recall curve (AUPRC) and the receiver operating characteristic curve (AUROCC), as well as optimal F1 score. When following RTOG guidelines, the models based on the nnU-Net auto segmentations (mean values: AUROCC, 0.71 ± 0.07; AUPRC, 0.42 ± 0.09; F1 score, 0.46 ± 0.08) significantly outperformed (P <.001) those based on the physician-delineated contours (mean values: AUROCC, 0.67 ± 0.07; AUPRC, 0.34 ± 0.08; F1 score, 0.36 ± 0.07). When spanning the entire bowel space, the performance of the autosegmentation models improved considerably (mean values: AUROCC, 0.87 ± 0.05; AUPRC, 0.70 ± 0.09; F1 score, 0.68 ± 0.09). Random forest models were superior at predicting acute grade ≥3 GI toxicity when based on RTOG-defined bowel bag autosegmentations rather than physician-delineated contours. Models based on autosegmentations spanning the entire bowel space show further considerable improvement in model performance. The results of this study should be further validated using an external data set. [ABSTRACT FROM AUTHOR]

Carbon dioxide (CO 2) is one of the most important greenhouse gases in the atmosphere, and carbon satellites play a vital role in monitoring its concentration. However, a single carbon satellite often has inadequate spatial coverage, resulting in numerous gaps. Utilizing the complementary advantages of multiple satellites in spatial coverage for high-coverage mapping of CO 2 may be an effective means. To this end, this study proposes a local random forest (LRF) model to generate a global high-coverage and high-precision column-averaged dry-air mole fraction of CO 2 (XCO 2) product from 2015 to 2021, which integrates multi-source data from OCO-2, OCO-3, GOSAT, and GOSAT-2. The results indicate that the LRF reconstructions agree well with the ground-based site observations, with RMSE, MAE, and R2 of 1.08 ppm, 0.82 ppm, and 0.96, respectively, outperforming the global random forest and artificial neural network models. Meanwhile, multi-source satellite fusion effectively improves the global XCO 2 coverage, with a maximum improvement rate of 122.2% compared to the single satellite of OCO-2. Based on the reconstructed dataset, regional and seasonal differences in the global XCO 2 distribution are observed, and an average growth rate of 2.18 ppm/year of global XCO 2 is revealed during 2015–2021. This study combines data from multi-source satellites to generate high-coverage global XCO 2 products from 2015 to 2021, which can be freely accessed from https://github.com/ch00en/Multi-source-satellite-XCO2-fusion-products.git and would hold great potential for carbon cycle research. [ABSTRACT FROM AUTHOR]

Gully erosion is a severe environmental issue that poses threats to agriculture, human safety, habitats, infrastructure, and soil integrity. Selecting the right machine learning model is vital for accurate Gully Erosion Susceptibility Mapping (GESM) due to varying environmental hazard performance. This study employed a comparative analysis of two machine learning techniques, Random Forest (RF) and Extreme Gradient Boost (XGBoost), to develop a highly precise GESM for the Silabati watershed (India). The analysis incorporated 24 controlling factors and examined a dataset of 460 sample points, with equal representation of gullies and non-gullies. Variance Inflation Factors (VIF) and Information Gain Ratio (IGR) techniques were applied to assess multicollinearity test among the controlling factors. Lithology, elevation, distance from the road, LULC, geomorphology, rainfall, drainage density, and coarse fragments emerged as crucial factors in determining GESM. Statistical tests, including the Kappa index, Root Mean Square Error (RMSE), Accuracy (ACC), Mean Absolute Error (MAE), Coefficient of Determination (R2), and Receiver Operating Characteristic (ROC), were employed to evaluate the RF and XGB models on training and testing data. Both models demonstrated strong performance, with the XGBoost and RF models achieving ROC values of 84.1% and 83.2%, respectively. The applied quantile classification method resulted in the creation of five distinct GESMs, categorized as very high (VH), high (H), moderate (M), low (L), and very low (VL). We found that the very high GESM areas of the watershed were 4.10% in the XGBoost model and 4.61%. in the RF model. Out of the 26 sub-watersheds, the results have identified five sub-watersheds as highly prioritized for sustainable management. Therefore, the present study provides an accurate gully erosion identification using advanced models, offering valuable insights for policymakers to proper management implementation. [ABSTRACT FROM AUTHOR]

Meningiomas display diverse biological traits and clinical behaviors, complicating patient outcome prediction. This heterogeneity, along with varying prognoses, underscores the need for a precise, personalized evaluation of postoperative outcomes. Data from the American College of Surgeons National Surgical Quality Improvement Program database identified patients who underwent intracranial meningioma resections from 2014 to 2020. We focused on 5 outcomes: prolonged LOS, nonhome discharges, 30-day readmissions, unplanned reoperations, and major complications. Six machine learning algorithms, including TabPFN, TabNet, XGBoost, LightGBM, Random Forest, and Logistic Regression, coupled with the Optuna optimization library for hyperparameter tuning, were tested. Models with the highest area under the receiver operating characteristic (AUROC) values were included in the web application. SHapley Additive exPlanations were used to evaluate the importance of predictor variables. Our analysis included 7000 patients. Of these patients, 1658 (23.7%) had prolonged LOS, 1266 (18.1%) had nonhome discharges, 573 (8.2%) had 30-day readmission, 253 (3.6%) had unplanned reoperation, and 888 (12.7%) had major complications. Performance evaluation indicated that the top-performing models for each outcome were the models built with LightGBM and Random Forest algorithms. The LightGBM models yielded AUROCs of 0.842 and 0.846 in predicting prolonged LOS and nonhome discharges, respectively. The Random Forest models yielded AUROCs of 0.717, 0.76, and 0.805 in predicting 30-day readmissions, unplanned reoperations, and major complications, respectively. The study successfully demonstrated the potential of machine learning models in predicting short-term adverse postoperative outcomes after meningioma resections. This approach represents a significant step forward in personalizing the information provided to meningioma patients. [ABSTRACT FROM AUTHOR]

Mechanical thrombectomy (MT) improves outcomes in patients with LVO but many still experience mortality or severe disability. We sought to develop machine learning (ML) models that predict 90-day outcomes after MT for LVO. Consecutive patients who underwent MT for LVO between 2015–2021 at a Comprehensive Stroke Center were reviewed. Outcomes included 90-day favorable functional status (mRS 0–2), severe disability (mRS 4–6), and mortality. ML models were trained for each outcome using prethrombectomy data (pre) and with thrombectomy data (post). Three hundred and fifty seven patients met the inclusion criteria. After model screening and hyperparameter tuning the top performing ML model for each outcome and timepoint was random forest (RF). Using only prethrombectomy features, the AUCs for the RF pre models were 0.73 (95% CI 0.62–0.85) for favorable functional status, 0.77 (95% CI 0.65–0.86) for severe disability, and 0.78 (95% CI 0.64–0.88) for mortality. All of these were better than a standard statistical model except for favorable functional status. Each RF model outperformed Pre, SPAN-100, THRIVE, and HIAT scores (P < 0.0001 for all). The most predictive features were premorbid mRS, age, and NIHSS. Incorporating MT data, the AUCs for the RF post models were 0.80 (95% CI 0.67–0.90) for favorable functional status, 0.82 (95% CI 0.69–0.91) for severe disability, and 0.71 (95% CI 0.55–0.84) for mortality. RF models accurately predicted 90-day outcomes after MT and performed better than standard statistical and clinical prediction models. [ABSTRACT FROM AUTHOR]

Intramedullary spinal cord tumors are challenging to resect, and their postoperative neurological outcomes are often difficult to predict, with few studies assessing this outcome. We reviewed the medical records of all patients surgically treated for Intramedullary spinal cord tumors at our multisite tertiary care institution (Mayo Clinic Arizona, Mayo Clinic Florida, Mayo Clinic Rochester) between June 2002 and May 2020. Variables that were significant in the univariate analyses were included in a multivariate logistic regression. "MissForest" operating on the Random Forest algorithm, was used for data imputation, and K-prototype was used for data clustering. Heatmaps were added to show correlations between postoperative neurological deficit and all other included variables. Shapley Additive exPlanations were implemented to understand each feature's importance. Our query resulted in 315 patients, with 160 meeting the inclusion criteria. There were 53 patients with astrocytoma, 66 with ependymoma, and 41 with hemangioblastoma. The mean age (standard deviation) was 42.3 (17.5), and 48.1% of patients were women (n = 77/160). Multivariate analysis revealed that pathologic grade >3 (OR = 1.55; CI = [0.67, 3.58], P = 0.046 predicted a new neurological deficit. Random Forest algorithm (supervised machine learning) found age, use of neuromonitoring, histology of the tumor, performing a midline myelotomy, and tumor location to be the most important predictors of new postoperative neurological deficits. Tumor grade/histology, age, use of neuromonitoring, and myelotomy type appeared to be most predictive of postoperative neurological deficits. These results can be used to better inform patients of perioperative risk. [ABSTRACT FROM AUTHOR]

Biorefining biological waste to produce eco-friendly fuels and by-products is essential in transitioning from non-renewable energies. However, the analysis of the processes in the laboratory necessitates a substantial investment of both time and money. The present study has developed machine learning (ML) models for evaluating biofuel products through glycerol biorefining for the first time. This study evaluates the performance of different ML algorithms, including recurrent neural network (RNN), random forest (RF), adaptive boosting (AdaBoost), Bayesian ridge (BR), and elastic net linear regression (ENLR). This research by machine learning algorithms can be created the formulas and the models for the hydrogen content (H 2) by the hydrogen production and time, the hydrogen production by the H 2 and time, OLR by PH and time, OLR by the hydrogen production and time, and the hydrogen production by OLR and time. Using the RNN for predicting the future of H 2 , hydrogen production, and OLR with the least error. The best R-Squared for the formulas is between 0.951 and 0.994 with the linear and the polynomial forms (by degrees 2, 3, 4). The best R-Squared for the models is between 0.998 and 0.999 with the linear form. MAEs for the formula and the model of H 2, respectively, are 2.475347126 and 0.46588143, and MAEs for the formula and the model of the hydrogen production, respectively, are 23.44120285 and 7.03283978. MAEs of OLR by PH and OLR by the hydrogen production for formulas and models are 2.095157 and 000001 by PH and 3.148667 and 0.000001 by the hydrogen production. MAEs for the formula and the model of the hydrogen production by OLR, respectively, are 19.025255 and 2.718604. • Development of machine learning models for evaluating biofuel products through glycerol biorefining. • Evaluation of different ML algorithms in predicting hydrogen production, H 2 content, and organic loading rate in the biorefining process. • Creation of formulas and models for H 2 content, hydrogen production, and OLR by various factors. • Use of RNN for predicting the future of H 2 content, hydrogen production, and OLR with the least error. • Best R-Squared for the models is between 0.998 and 0.999 with the linear form. [ABSTRACT FROM AUTHOR]

The demand for high-performance displays is driving the development of glass substrates with extremely low strain variation during manufacturing. To achieve the necessary balance of physical properties to resist strain, the traditional compositional design methods for glass substrates like trial-and-error and classical computational techniques should be optimized. As an alternative approach, machine learning (ML) algorithms have emerged for designing new glass compositions. In this work, we conduct ML research focused on the compositional design for high-performance glass substrates. We employ three ML algorithms i.e., Random Forest (RF), Classification And Regression Tree (CART) and k-Nearest Neighbors (k-NN) to predict five physical properties that are key to the performance of alkaline-free aluminosilicate glass substrates. By using only small-to-medium dataset sizes, our model reaches a high coefficient of determination of 0.9879. Furthermore, our model achieves an accurate generalized prediction of 50 experimental data and enables the prediction and design of glass substrate compositions with advanced comprehensive properties. [ABSTRACT FROM AUTHOR]

In this work, we study the numerical solution of inverse eigenvalue problems from a machine learning perspective. Two different problems are considered: the inverse Sturm-Liouville eigenvalue problem for symmetric potentials and the inverse transmission eigenvalue problem for spherically symmetric refractive indices. Firstly, we solve the corresponding direct problems to produce the required eigenvalues datasets in order to train the machine learning algorithms. Next, we consider several examples of inverse problems and compare the performance of each model to predict the unknown potentials and refractive indices respectively, from a given small set of the lowest eigenvalues. The supervised regression models we use are k-Nearest Neighbours, Random Forests and Multi-Layer Perceptron. Our experiments show that these machine learning methods, under appropriate tuning on their parameters, can numerically solve the examined inverse eigenvalue problems. [ABSTRACT FROM AUTHOR]

The total electron content (TEC) is a crucial parameter for ionosphere monitoring, and the development of accurate TEC estimation and prediction models is significant for high-precision positioning. However, establishing a high-precision ionospheric model that can effectively improve prediction accuracy remains a challenging task. We propose a method that combines random forest with a Bi-LSTM neural network to establish a high-precision ionospheric prediction model. The random forest algorithm is used for regression analysis and to estimate the variable importance of input parameters. We use observation data from the Crustal Movement Observation Network of China and International GNSS Services while estimating the relative importance of 14 input variables at different latitudes. The results demonstrate that our proposed model achieves more efficient and accurate predictions, with a correlation coefficient of 0.96, indicating significant improvement in accuracy compared to traditional RNN and LSTM models. Overall, the Bi-LSTM forecast model based on random forest can effectively capture the temporal and spatial variation characteristics of ionospheric TEC in China. This enables the model to provide accurate ionospheric information for positioning users, thereby improving the precision of positioning applications. [ABSTRACT FROM AUTHOR]

Depression is a major challenge in the global healthy aging process, and exploring the key factors of depression in urban and rural older adults is essential for differentiated and precise interventions. To explore the urban-rural differences and key influencing factors of depressive symptoms among Chinese older adults. The data of 5267 older adults were obtained from the China Health and Retirement Longitudinal Survey (CHARLS2018). Random forest model and logistic regression were used to analyze the key factors influencing depressive symptoms 19 variables. The detection rate of depressive symptoms in older adults was 31.0 %, with 22.3 % and 34.8 % in urban and rural areas, respectively. Education, self-rated health, self-rated pain, and self-rated vision were common factors. Physical activity (OR = 0.716 for Moderate PA), social activity (OR = 0.671 for social activity), and self-rated hearing (OR = 0.602 for good) were key factors specific to urban older adults, and alcohol consumption (OR = 0.716 for drinking more than once a month) and marital status (OR = 0.689 for cohabitation) were key factors specific to depressive symptoms in rural older adults (all P < 0.05). Cross-sectional data cannot reflect dynamic processes among variables; The cultural background might affect the cross-cultural validity of the study. The key factors of depressive symptoms among older adults in urban and rural areas differed, which provides priority and references for differential prevention and precise intervention of depressive symptoms to promote the process of healthy aging. • Education, self-rated health, pain, and vision are common factors that influence depressive symptoms in older adults. • Physical activity, social activity, and self-rated hearing were key factors specific to urban older adults. • Alcohol consumption and marital status were key factors specific to rural older adults. • The combination of random forest and logistic regression achieves the importance measurement and analysis of factors. [ABSTRACT FROM AUTHOR]

The polysialyltransferases ST8SIA2 and ST8SIA4 and their product, polysialic acid (polySia), are known to be related to cancers and mental disorders. ST8SIA2 and ST8SIA4 have conserved amino acid (AA) sequence motifs essential for the synthesis of the polySia structures on the neural cell adhesion molecule. To search for a new motif in the polysialyltransferases, we adopted the in silico Individual Meta Random Forest program that can predict disease-related AA substitutions. The Individual Meta Random Forest program predicted a new eight-amino-acids sequence motif consisting of highly pathogenic AA residues, thus designated as the pathogenic (P) motif. A series of alanine point mutation experiments in the pathogenic motif (P motif) showed that most P motif mutants lost the polysialylation activity without changing the proper enzyme expression levels or localization in the Golgi. In addition, we evaluated the enzyme stability of the P motif mutants using newly established calculations of mutation energy, demonstrating that the subtle change of the conformational energy regulates the activity. In the AlphaFold2 model, we found that the P motif was a buried ß-strand underneath the known surface motifs unique to ST8SIA2 and ST8SIA4. Taken together, the P motif is a novel buried ß-strand that regulates the full activity of polysialyltransferases from the inside of the molecule. [ABSTRACT FROM AUTHOR]

Prediction of treatment resistance in schizophrenia (TRS) would be helpful to reduce the duration of ineffective treatment and avoid delays in clozapine initiation. We applied machine learning to identify clinical, sociodemographic, familial, and environmental variables that are associated with TRS and could potentially predict TRS in the future. Baseline and follow-up data on trait(-like) variables from the Genetic Risk and Outcome of Psychosis (GROUP) study were used. For the main analysis, we selected patients with non-affective psychotic disorders who met TRS (n = 200) or antipsychotic-responsive criteria (n = 423) throughout the study. For a sensitivity analysis, we only selected patients who met TRS (n = 76) or antipsychotic-responsive criteria (n = 123) at follow-up but not at baseline. Random forest models were trained to predict TRS in both datasets. SHapley Additive exPlanation values were used to examine the variables' contributions to the prediction. Premorbid functioning, age at onset, and educational degree were most consistently associated with TRS across both analyses. Marital status, current household, intelligence quotient, number of moves, and family loading score for substance abuse also consistently contributed to the prediction of TRS in the main or sensitivity analysis. The diagnostic performance of our models was modest (area under the curve: 0.66–0.69). We demonstrate that various clinical, sociodemographic, familial, and environmental variables are associated with TRS. Our models only showed modest performance in predicting TRS. Prospective large multi-centre studies are needed to validate our findings and investigate whether the model's performance can be improved by adding data from different modalities. [ABSTRACT FROM AUTHOR]

To explore the clinical value of a clinical radiomics model nomogram based on magnetic resonance imaging (MRI) for preoperative meningioma grading. We collected retrospectively 544 patients with pathological diagnosis of meningiomas were categorized into training (n = 380) and validation (n = 164) groups at the ratio of 7∶ 3. There were 3,376 radiomics features extracted from T2WI and T1C by shukun technology platform after manual segmentation using an independent blind method by two radiologists. The Selectpercentile and Lasso are used to filter the most strongly correlated features. Random forest (RF) radiomics model and clinical radiomics model nomogram were constructed respectively. The calibration, discrimination, and clinical validity were evaluated by using the calibration curve and decision analysis curve (DCA). The RF radiomics model based on T1C and T2WI was the most effective to predict meningioma grade before surgery among the six different classifiers. The predictive ability of clinical radiomics model was slightly higher than that of RF model alone. The AUC, SEN, SPE, and ACC of the training set were 0.949, 0.976, 0.785, and 0.826, and the AUC, SEN, SPE, and ACC of the validation set were 0.838, 0.829, 0.783, and 0.793, respectively. The calibration curve and Hosmer-Lemeshow test showed the predictive probability of the fusion model was similar to the actual differentiated LGM and HGM. The analysis of the decision curve showed that the clinical radiomics model could obtain the best clinical net profit. The clinical radiomics model nomogram based on T1C and T2WI has high accuracy and sensitivity for predicting meningioma grade. [ABSTRACT FROM AUTHOR]

Data-driven modeling method is an effective method for large-scale refinery industrial units. However, due to the long operational cycle, large number of measurement points, complex spatial and temporal correlations and dynamic switching of operation modes of refinery industrial units, utilizing a single data-driven model for process modeling may not be optimal. To tackle with the issues above, a self-attention model based on feature selection and pattern classification (FS-PC-SA) is develop. This approach uses Random Forest algorithm for feature selection and extract different operating modes from industrial data through K-means clustering algorithm. To capture complex spatial and temporal correlations in refinery units, sub-models based on self-attention for each mode are built for prediction. Finally, the method is used to predict the gasoline product yield in fluid catalytic cracking process and wax oil product yield in ebullated bed residue hydrocracking process, which demonstrates that the proposed method is better than other traditional methods. • The pattern classification of operating conditions solves the problem of low prediction accuracy of a single model. • The self-attention mechanism is used to capture complex data relationships. • Multi-mode industrial modeling framework is established and applied to the actual industrial process. [ABSTRACT FROM AUTHOR]

• Pulmonary embolism (PE) patients who are low risk benefit from early discharge. • A machine learning model is developed to predict short and long term mortality in PE. • This model improved on standard clinical risk stratification for PE patients. • External validation is required to determine generalizability of the model. Pulmonary embolism (PE) is a severe condition that causes significant mortality and morbidity. Due to its acute nature, scores have been developed to stratify patients at high risk of 30-day mortality. Here we develop a machine-learning based score to predict 30-day, 90-day, and 365-day mortality in PE patients. The Birmingham and Black Country Venous Thromboembolism registry (BBC-VTE) of 2183 venous thromboembolism patients is used. Random forests were trained on a 70% training cohort and tested against 30% held-out set. The outcomes of interest were 30-day, 90-day, and 365-day mortality. These were compared to the pulmonary embolism severity index (PESI) and simplified pulmonary embolism severity index (sPESI). Shapley values were used to determine important predictors. Oral anticoagulation at discharge was also investigated as a predictor of mortality. The machine learning risk score predicted 30-day mortality with AUC 0.71 [95% CI: 0.63 - 0.78] compared to the sPESI AUC of 0.65 [95% CI: 0.57 - 0.73] and PESI AUC of 0.64 [95% CI: 0.56 – 0.72]. 90-day mortality and 365-day mortality were predicted with an AUC of 0.74 and 0.73 respectively. High counts of neutrophils, white blood cell counts, and c-reactive protein and low counts of haemoglobin were important for 30-day mortality prediction but progressively lost importance with time. Older age was an important predictor of high risk throughout. Machine learning algorithms have improved on standard clinical risk stratification for PE patients. External cohort validation is required before incorporation into clinical workflows. [ABSTRACT FROM AUTHOR]

• Difference of 21 metabolites and 38 lipids in hypertensive IS and healthy subjects. • Difference of 18 metabolites and 31 lipids in hypertensive IS and hypertension subjects. • Two 7 biomarkers panels distinguished hypertensive IS from healthy and hypertension. • Medication use did not affect the performance of the diagnostic model. Hypertensive individuals are at a high risk of stroke, and thus, prevention of stroke in hypertensive patients is essential. Metabolomics and lipidomics can be used to identify diagnostic biomarkers and conduct early assessments of stroke risk in hypertensive populations. In this study, serum samples were collected from 30 hypertensive ischemic stroke (IS), 30 matched hypertensive and 30 matched healthy participants. Metabolomics and lipidomics analyses were conducted via liquid chromatography–tandem mass spectrometry, and the data were analyzed using multivariate and univariate statistical methods. A random forest algorithm and binary logistic regression were used to screen the biomarkers and establish diagnostic model. We detected 21 differential metabolites and 38 differential lipids between the hypertensive IS and healthy group. Moreover, we found 18 differential metabolites and 31 differential lipids between the hypertensive IS and hypertension group. In particular, the following seven metabolites or lipids distinguished the hypertensive IS from the healthy group: 4-hydroxyphenylpyruvic acid, cafestol, phosphatidylethanolamine (PE) (18:0p/18:2), PE (16:0e/20:4), (O -acyI)-1-hydroxy fatty acid (36:3), PE (16:0p/20:3) and PE (18:1p/18:2) (rep). The following seven biomarkers distinguished the hypertensive IS from the hypertension group: diglyceride (DG) (20:1/18:2), PE (18:0p/18:2), PE (16:0e/22:5), phosphatidylcholine (40:7), dimethylphosphatidylethanolamine (50:3), DG (18:1/18:2), and 4-hydroxyphenylpyruvic acid. The aforementioned panels had good diagnostic and predictive ability for hypertensive IS. Our study determines the metabolomic and lipidomic profiles of hypertensive IS patients and thereby identifies potential biomarkers of the presence of IS in hypertensive populations. [ABSTRACT FROM AUTHOR]

• The potential of optical and SAR data was explored for estimating grassland height. • The RF, GBRT, and CART models were used to estimate grassland height. • GBRT model performed than RF and CART models for grassland height estimation. • The optical remote sensing data performed better than SAR remote sensing data. • Coupling optical and SAR data can effectively improve the estimation accuracy. Grassland height is an important indicator used to evaluate ecological environments in grasslands. The study explored the potential of optical and SAR (synthetic aperture radar) remote sensing data in estimating grassland height with machine learning methods and constructed the grassland height remote sensing inversion model. Then the mean grassland height in August 2015 was estimated with the inversion model in Inner Mongolia Autonomous Region, China. Among classification and regression tree (CART), random forest (RF), and gradient boosting regression tree (GBRT) models, the GBRT model had the highest estimation accuracy. There was little correlation between SAR data and grassland height, and SAR data produced poor accuracy for grassland height estimation. The grassland height could be better estimated by optical remote sensing data. The best estimation accuracy (GBRT model: for training data: R2 = 0.71, RMSE = 3.58 cm, P < 0.01; for test data: R2 = 0.58, RMSE = 3.94 cm, P < 0.01) was achieved by the combination of optical and SAR remote sensing data. However, SAR data played an auxiliary role, and the estimation of grassland height was mainly realized by optical data. The average vegetation height of the grasslands in Inner Mongolia in August 2015 was 19.15 cm, gradually decreasing from northeast to southwest. The study proposed a high-precision method for estimating grassland height, which provided a basis for studying grassland environments and conditions. [ABSTRACT FROM AUTHOR]

• Study using Sentinel-2 for inland Arctic bathymetry retrieval was conducted. • Bathymetry was modelled using Google Earth Engine and machine learning. • Machine learning algorithm efficacy and robustness was carefully investigated. • The Random Forest algorithm produced the most accurate results. Lakes and ponds are extensive features throughout the circumpolar region, spanning a broad range of environmental conditions which controls their hydro-ecological processes and spatiotemporal distribution. The physical characteristics of these freshwater ecosystems, including their extent and depth, are particularly responsive to climatic conditions. Thus, having the ability to efficiently map and monitor these elements is crucial as the climate continues to warm, especially over large spatial extents. In this study, satellite derived bathymetry (SDB) methods were implemented to model regional inland Arctic water depths within the open-source and cloud-based Google Earth Engine (GEE) platform. High-resolution (10 m) spectral reflectance data from Sentinel-2 was used as covariates in non-parametric and non-linear machine learning (ML) models, namely random forest (RF), support vector regression (SVR), and classification and regression trees (CART). These three ML models were also compared to a more classical and parametric multiple linear regression (MLR) model. All models were calibrated using in situ bathymetric data collected near the Toolik Field Station, in the Alaskan Arctic tundra. With such a large sample repository available, the effects of training set size, noise, and multicollinearity on model performance was comprehensively investigated. Results clearly demonstrated the superior efficacy and robustness of the RF algorithm for predicting water depths, achieving a best coefficient of determination (R2) of 0.74, mean absolute error (MAE) of 2.12 m, and root mean square error (RMSE) of 3.04 m. Overall, this study highlights the potential of advanced and flexible ML algorithms within the GEE, and demonstrates the capabilities of the polar-orbiting Copernicus Sentinel-2 satellite for this Arctic application. In the future, this ML-driven methodology can be applied using GEE's cloud infrastructure to produce updated, cost-effective, and accurate bathymetry maps over large expanses of the Arctic tundra. [ABSTRACT FROM AUTHOR]