Predicting Palustrine Wetland Probability Using Random Forest Machine Learning and Digital Elevation Data-Derived Terrain Variables

The probabilistic mapping of landslide occurrence at a high spatial resolution and over a large geographic extent is explored using random forests (RF) machine learning; light detection and ranging (LiDAR)-derived terrain variables; additional variables relating to lithology, soils, distance to roads and streams and cost distance to roads and streams; and training data interpreted from high spatial resolution LiDAR-derivatives. Using a large training set and all predictor variables, an area under the receiver operating characteristic (ROC) curve (AUC) of 0.946 is obtained. Our findings highlight the value of a large training dataset, the incorporation of a variety of terrain variables and the use of variable window sizes to characterize the landscape at different spatial scales. We also document important variables for mapping slope failures. Our results suggest that feature selection is not required to improve the RF modeling results and that incorporating multiple models using different pseudo absence samples is not necessary. From our findings and based on a review of prior studies, we make recommendations for high spatial resolution, large-area slope failure probabilistic mapping.

Culicoides biting midges exhibit a global spatial distribution and are the main vectors of several viruses of veterinary importance, including bluetongue (BT) and African horse sickness (AHS). Many environmental and anthropological factors contribute to their ability to live in a variety of habitats, which have the potential to change over the years as the climate changes. Therefore, as new habitats emerge, the risk for new introductions of these diseases of interest to occur increases. The aim of this study was to model distributions for two primary vectors for BT and AHS (Culicoides imicola and Culicoides bolitinos) using random forest (RF) machine learning and explore the relative importance of environmental and anthropological factors in a region of South Africa with frequent AHS and BT outbreaks. Culicoides capture data were collected between 1996 and 2022 across 171 different capture locations in the Western Cape. Predictor variables included climate-related variables (temperature, precipitation, humidity), environment-related variables (normalised difference vegetation index-NDVI, soil moisture) and farm-related variables (livestock densities). Random forest (RF) models were developed to explore the spatial distributions of C. imicola, C. bolitinos and a merged species map, where both competent vectors were combined. The maps were then compared to interpolation maps using the same capture data as well as historical locations of BT and AHS outbreaks. Overall, the RF models performed well with 75.02%, 61.6% and 74.01% variance explained for C. imicola, C. bolitinos and merged species models respectively. Cattle density was the most important predictor for C. imicola and water vapour pressure the most important for C. bolitinos. Compared to interpolation maps, the RF models had higher predictive power throughout most of the year when species were modelled individually; however, when merged, the interpolation maps performed better in all seasons except winter. Finally, midge densities did not show any conclusive correlation with BT or AHS outbreaks. This study yielded novel insight into the spatial abundance and drivers of abundance of competent vectors of BT and AHS. It also provided valuable data to inform mathematical models exploring disease outbreaks so that Culicoides-transmitted diseases in South Africa can be further analysed.

The article analyzes current trends in the use of artificial intelligence, cognitive technologies, Big Data and other innovations in order to predict the stability of the Russian banking system. The current trends in the development of banking innovations in the context of digitalization of the economy are summarized. Among the research methods used, the Random Forest artificial intelligence system, formed in the Google Collab cloud service in Python, should be noted. A neural network model for predicting the dynamics of bank profits based on the use of the Random Forest (RF – random forest) machine learning model is proposed. The results of the study indicate that, with the average value of the profit of banks for the period 2010-2021. 1126.0416 billion rubles, the mean absolute error (Mean Absolute Error) amounted to 129.96666 billion rubles, or 11.5%, which can be regarded as a good result in conditions of market uncertainty. The analysis showed that current trends are global in nature and lead to qualitative changes in the flow of business processes between consumers of banking services and credit institutions.
 Purpose: is to form a model that predicts the value of the target variable – the profit of the banking system using the Random Forest deep learning model, which is an ensemble of “decision trees”. The scientific novelty is to put forward and prove the hypothesis that with the help of the formed artificial intelligence system “Random Forest”, a profit forecast for commercial banks of the Russian Federation can be obtained.
 Method or methodology of the work. The work used such research methods as: monographic, analytical, artificial intelligence system – “Random Forest”, as well as comparative analysis. Calculations were performed in XL tables, Google Collab cloud service in Python.
 Results. The developed AI-system “Random Forest”, designed to predict the profits of banks, is presented. It is of practical importance to use the results of the Random Forest neural network to predict the development of credit institutions, which makes it possible to solve a major national economic problem - ensuring the stability of the banking system of the Russian Federation. A neural network model for predicting the dynamics of bank profits based on the use of the Random Forest (RF) machine learning model is proposed, which is a machine learning algorithm that consists in using an ensemble (set) of decision trees.
 The scope of the results: the financial sector, the banking sector, forecasting and ensuring the sustainable activity of credit institutions in the Russian Federation.

Abstract. We investigated the potential capability of the random forest (RF) machine learning (ML) model to estimate snow depth in this work. Four combinations composed of critical predictor variables were used to train the RF model. Then, we utilized three validation datasets from out-of-bag (OOB) samples, a temporal subset, and a spatiotemporal subset to verify the fitted RF algorithms. The results indicated the following: (1) the accuracy of the RF model is greatly influenced by geographic location, elevation, and land cover fractions; (2) however, the redundant predictor variables (if highly correlated) slightly affect the RF model; and (3) the fitted RF algorithms perform better on temporal than spatial scales, with unbiased root-mean-square errors (RMSEs) of ∼4.4 and ∼7.3 cm, respectively. Finally, we used the fitted RF2 algorithm to retrieve a consistent 32-year daily snow depth dataset from 1987 to 2018. This product was evaluated against the independent station observations during the period 1987–2018. The mean unbiased RMSE and bias were 7.1 and −0.05 cm, respectively, indicating better performance than that of the former snow depth dataset (8.4 and −1.20 cm) from the Environmental and Ecological Science Data Center for West China (WESTDC). Although the RF product was superior to the WESTDC dataset, it still underestimated deep snow cover (>20 cm), with biases of −10.4, −8.9, and −34.1 cm for northeast China (NEC), northern Xinjiang (XJ), and the Qinghai–Tibetan Plateau (QTP), respectively. Additionally, the long-term snow depth datasets (station observations, RF estimates, and WESTDC product) were analyzed in terms of temporal and spatial variations over China. On a temporal scale, the ground truth snow depth presented a significant increasing trend from 1987 to 2018, especially in NEC. However, the RF and WESTDC products displayed no significant changing trends except on the QTP. The WESTDC product presented a significant decreasing trend on the QTP, with a correlation coefficient of −0.55, whereas there were no significant trends for ground truth observations and the RF product. For the spatial characteristics, similar trend patterns were observed for RF and WESTDC products over China. These characteristics presented significant decreasing trends in most areas and a significant increasing trend in central NEC.

BackgroundCalcific aortic valve disease (CAVD) is one of the most prevalent valvular diseases and is the second most common cause for cardiac surgery. However, the mechanism of CAVD remains unclear. This study aimed to investigate the role of pyroptosis-related genes in CAVD by performing comprehensive bioinformatics analysis.MethodsThree microarray datasets (GSE51472, GSE12644 and GSE83453) and one RNA sequencing dataset (GSE153555) were obtained from the Gene Expression Omnibus (GEO) database. Pyroptosis-related differentially expressed genes (DEGs) were identified between the calcified and the normal valve samples. LASSO regression and random forest (RF) machine learning analyses were performed to identify pyroptosis-related DEGs with diagnostic value. A diagnostic model was constructed with the diagnostic candidate pyroptosis-related DEGs. Receiver operating characteristic (ROC) curve analysis was performed to estimate the diagnostic performances of the diagnostic model and the individual diagnostic candidate genes in the training and validation cohorts. CIBERSORT analysis was performed to estimate the differences in the infiltration of the immune cell types. Pearson correlation analysis was used to investigate associations between the diagnostic biomarkers and the immune cell types. Immunohistochemistry was used to validate protein concentration.ResultsWe identified 805 DEGs, including 319 down-regulated genes and 486 up-regulated genes. These DEGs were mainly enriched in pathways related to the inflammatory responses. Subsequently, we identified 17 pyroptosis-related DEGs by comparing the 805 DEGs with the 223 pyroptosis-related genes. LASSO regression and RF algorithm analyses identified three CAVD diagnostic candidate genes (TREM1, TNFRSF11B, and PGF), which were significantly upregulated in the CAVD tissue samples. A diagnostic model was constructed with these 3 diagnostic candidate genes. The diagnostic model and the 3 diagnostic candidate genes showed good diagnostic performances with AUC values >0.75 in both the training and the validation cohorts based on the ROC curve analyses. CIBERSORT analyses demonstrated positive correlation between the proportion of M0 macrophages in the valve tissues and the expression levels of TREM1, TNFRSF11B, and PGF.ConclusionThree pyroptosis-related genes (TREM1, TNFRSF11B and PGF) were identified as diagnostic biomarkers for CAVD. These pyroptosis genes and the pro-inflammatory microenvironment in the calcified valve tissues are potential therapeutic targets for alleviating CAVD.

Citrus greening disease (Huanglongbing-HLB) is considered the most destructive citrus disease worldwide. Of the three species of Candidatus liberibacter associated with HLB, two have been recently reported in Latin America. The first report of HLB in Colombia was in March 2016. In this paper, a dataset was extracted for six departments in the northern zone of Colombia, where has been previously reported, applying image georeferencing with QGIS Software. Preliminary Random Forest and K-Nearest Neighbors (KNN) machine learning models were used in order to test and validate the obtained results, for disease monitoring and HLB incidence prediction. The performance of both models was also compared, obtaining a 100% AUC value with Random Forest model.

Background/Objectives: Acute respiratory distress syndrome (ARDS) represents a critical complication in polytrauma patients, characterized by diffuse lung inflammation and bilateral pulmonary infiltrates with mortality rates reaching 45% in intensive care units (ICU). The heterogeneous nature of ARDS and complex clinical presentation in severely injured patients poses substantial diagnostic challenges, necessitating early prediction tools to guide timely interventions. Machine learning (ML) algorithms have emerged as promising approaches for clinical decision support, demonstrating superior performance compared to traditional scoring systems in capturing complex patterns within high-dimensional medical data. Based on the identified research gaps in early ARDS prediction for polytrauma populations, our study aimed to: (i) develop a balanced random forest (BRF) ML model for early ARDS prediction in critically ill polytrauma patients, (ii) identify the most predictive clinical features using ANOVA-based feature selection, and (iii) evaluate model performance using comprehensive metrics addressing class imbalance challenges. Methods: This retrospective cohort study analyzed 407 polytrauma patients admitted to the ICU of the Center of Traumatology and Major Burns of Ben Arous, Tunisia, between 2017 and 2021. We implemented a comprehensive ML pipeline that incorporates Tomek Links undersampling, ANOVA F-test feature selection for the top 10 predictive variables, and SMOTE oversampling with a conservative sampling rate of 0.3. The BRF classifier was trained with class weighting and evaluated using stratified 5-fold cross-validation. Performance metrics included AUROC, PR-AUC, sensitivity, specificity, F1-score, and Matthews correlation coefficient. Results: Among 407 patients, 43 developed ARDS according to the Berlin definition, representing a 10.57% incidence. The BRF model demonstrated exceptional predictive performance with an AUROC of 0.98, a sensitivity of 0.91, a specificity of 0.80, an F1-score of 0.84, and an MCC of 0.70. Precision-recall AUC reached 0.86, demonstrating robust performance despite class imbalance. During stratified cross-validation, AUROC values ranged from 0.93 to 0.99 across folds, indicating consistent model stability. The top 10 selected features included procalcitonin, PaO2 at ICU admission, 24-h pH, massive transfusion, total fluid resuscitation, presence of pneumothorax, alveolar hemorrhage, pulmonary contusion, hemothorax, and flail chest injury. Conclusions: Our BRF model provides a robust, clinically applicable tool for early prediction of ARDS in polytrauma patients using readily available clinical parameters. The comprehensive two-step resampling approach, combined with ANOVA-based feature selection, successfully addressed class imbalance while maintaining high predictive accuracy. These findings support integrating ML approaches into critical care decision-making to improve patient outcomes and resource allocation. External validation in diverse populations remains essential for confirming generalizability and clinical implementation.

Land Use and Land Cover (LULC) dynamics have a major impact on environmental sustainability, re-source management, and urban development. Effective decision-making depends on correct forecasting of these changes. This research predicts LULC changes with the Random Forest (RF) machine learning tech-nique using satellite-derived data from the National Remote Sensing Centre (NRSC) for the years from 2005 to 2023. The dataset includes various LULC categories such as built-up land, agricultural lands, plan-tation/orchards, forests, wetlands, grasslands, wastelands, and water bodies. This study focuses on an area of Tamilnadu, India, covered by NRSC’s LULC datasets from 2005 to 2023. The Random Forest model was first trained on data for the years from 2005 to 2017 to predict LULC for the next consecutive years 2018–2023, with validated against actual LULC values for Tamilnadu, India from the year 2018 to 2023, achieving high accuracy with a correlation coefficient (R > 0.97 in later years) and decreasing Mean Abso-lute Error. Based on the complete historical dataset from 2005 to 2023, the trained model is then applied to predict LULC changes for the years from 2024 to 2028. The results indicate a significant increase in urban development of 0.001 hectare annually and a consistent decrease in double/triple cropping areas of ap-proximately 1.5 hectare between 2024 and 2028, stable but slightly declining forest cover, and water body spread oscillations. The steady increase in built-up land underscores the importance of controlled urban expansion and the decline in double/triple cropping areas calls for policies that support sustainable farm-ing practices. With forest cover slightly declining, policymakers should strengthen conservation initia-tives, afforestation efforts, and enforce stricter land-use regulations to prevent degradation. The oscillating spread of water bodies highlights the need for improved watershed management strategies.

The management and monitoring of land use in geothermal fields are crucial for the sustainable utilization of water resources, as well as for striking a balance between the production of renewable energy and the preservation of the environment. This study primarily compared Support Vector Machine (SVM) and Random Forest (RF) machine learning methods, using satellite imagery from Landsat 8 and Sentinel 2 between 2021 and 2023, to monitor land use in the Patuha geothermal area. The objective is to improve sustainable water management practices by accurately categorizing different land cover types. This comparative analysis assessed the efficacy of these techniques in upholding water sustainability in geothermal regions. This study examined the application of SVM and RF machine learning techniques, with particular emphasis on parameter refinement and model assessment, to enhance land use classification accuracy. By employing Kernlab and e1071 for algorithm comparison, the research sought to produce a precise Land Use Model Map, which underscores the significance of advanced analytical techniques in environmental management. This approach was of utmost importance in improving land use monitoring and reinforcing sustainable practices. The comparative evaluation of SVM and RF methods for land use classification demonstrates the superiority of RF in terms of accuracy, stability, and precision, particularly in intricate urban settings, hence establishing it as the preferred model for tasks demanding high reliability. The application of SVM and RF for monitoring land use in geothermal areas is in alignment with Sustainable Development Goals (SDGs) 6 and 15, as it fosters sustainable water management and the conservation of ecosystems. Doi: 10.28991/HEF-2024-05-02-06 Full Text: PDF

Seasonal fires in northern Thailand are a persistent environmental and public health concern, yet existing fire probability mapping approaches in Thailand rely heavily on subjective multi-criteria analysis (MCA) methods and temporally static data aggregation methods. To address these limitations, we present a flexible, replicable, and operationally viable seasonal fire probability mapping methodology using a Random Forest (RF) machine learning model in the Google Earth Engine (GEE) platform. We trained the model on historical fire occurrence and fire predictor layers from 2016–2023 and applied it to 2024 conditions to generate a probabilistic fire prediction. Our novel approach improves upon existing operational methods and scientific literature in several ways. It uses a more representative sample design which is agnostic to the burn history of fire presences and absences, pairs fire and fire predictor data from each year to account for interannual variation in conditions, empirically refines the most influential fire predictors from a comprehensive set of predictors, and provides a reproducible and accessible framework using GEE. Predictor variables include both socioeconomic and environmental drivers of fire, such as topography, fuels, potential fire behavior, forest type, vegetation characteristics, climate, water availability, crop type, recent burn history, and human influence and accessibility. The model achieves an Area Under the Curve (AUC) of 0.841 when applied to 2016–2023 data and 0.848 when applied to 2024 data, indicating strong discriminatory power despite the additional spatial and temporal variability introduced by our sample design. The highest fire probabilities emerge in forested and agricultural areas at mid elevations and near human settlements and roads, which aligns well with the known anthropogenic drivers of fire in Thailand. Distinct areas of model uncertainty are also apparent in cropland and forests which are only burned intermittently, highlighting the importance of accounting for localized burning cycles. Variable importance analysis using the Gini Impurity Index identifies both natural and anthropogenic predictors as key and nearly equally important predictors of fire, including certain forest and crop types, vegetation characteristics, topography, climate, human influence and accessibility, water availability, and recent burn history. Our findings demonstrate the heavy influence of data preprocessing and model design choices on model results. The model outputs are provided as interpretable probability maps and the methods can be adapted to future years or augmented with local datasets. Our methodology presents a scalable advancement in wildfire probability mapping with machine learning and open-source tools, particularly for data-constrained landscapes. It will support Thailand’s fire managers in proactive fire response and planning and also inform broader regional fire risk assessment efforts.

ABSTRACTSentinel-1A synthetic aperture radar (SAR) data present an opportunity for acquiring crop information without restrictions caused by weather and illumination conditions, at a spatial resolution appropriate for individual rice fields and a temporal resolution sufficient to capture the growth profiles of different crop species. This study investigated the use of multi-temporal Sentinel-1A SAR data and Landsat-derived normalized difference vegetation index (NDVI) data to map the spatial distribution of paddy rice fields across parts of the Sanjiang plain, in northeast China. The satellite sensor data were acquired throughout the rice crop-growing season (May–October). A co-registered set of 10 dual polarization (VH/VV) SAR and NDVI images depicting crop phenological development were used as inputs to Support Vector Machine (SVM) and Random Forest (RF) machine learning classification algorithms in order to map paddy rice fields. The results showed a significant increase in overall classification when the NDVI time-series data were integrated with the various combinations of multi-temporal polarization channels (i.e. VH, VV, and VH/VV). The highest classification accuracies overall (95.2%) and for paddy rice (96.7%) were generated using the RF algorithm applied to combined multi-temporal VH polarization and NDVI data. The SVM classifier was most effective when applied to the dual polarization (i.e. VH and VV) SAR data alone and this generated overall and paddy rice classification accuracies of 91.6% and 82.5%, respectively. The results demonstrate the practicality of implementing RF or SVM machine learning algorithms to produce 10 m spatial resolution maps of paddy rice fields with limited ground data using a combination of multi-temporal SAR and NDVI data, where available, or SAR data alone. The methodological framework developed in this study is apposite for large-scale implementation across China and other major rice-growing regions of the world.

Fluvial hazards of river mobility and flooding are often problematic for road infrastructure and need to be considered in the planning process. The extent of river and road infrastructure networks and their tendency to be close to each other creates a need to be able to identify the most dangerous areas quickly and cost‐effectively. In this study, we propose a novel methodology using random forest (RF) machine learning methods to provide easily interpretable fine‐scale fluvial hazard predictions for large river systems. The tools developed provide predictions for three models: presence of flooding (PFM), presence of mobility (PMM) and type of erosion model (TEM, lateral migration, or incision) at reference points every 100 m along the fluvial network of three watersheds within the province of Quebec, Canada. The RF models use variables focused on river conditions and hydrogeomorphological processes such as confinement, sinuosity, and upstream slope. Training/validation data included field observations, results from hydraulic and erosion models, government infrastructure databases, and hydro‐ geomorphological assessments using 1‐m DEM and satellite/historical imagery. A total of 1807 reference points were classified for flooding, 1542 for mobility, and 847 for the type of erosion out of the 11,452 reference points for the 1145 km of rivers included in the study. These were divided into training (75%) and validation (25%) datasets, with the training dataset used to train supervised RF models. The validation dataset indicated the models were capable of accurately predicting the potential for fluvial hazards to occur, with precision results for the three models ranging from 83% to 94% of points accurately predicted. The results of this study suggest that RF models are a cost‐effective tool to quickly evaluate the potential for fluvial hazards to occur at the watershed scale.

Characterization of surface–groundwater interaction is an increasingly useful skill for riverine ecologists and water resource managers interested in the dynamics of water, nutrient, and micro‐organism exchange at the reach scale, as it can be used to better represent point‐scale processes within larger catchment‐scale models. This study describes a method for predicting the nature of reach‐scale surface–groundwater interaction, using the random forest (RF) machine learning technique with national‐scale geology, hydrology, and land use data. Observed stream flow depletion and accretion surveys from riparian areas and spring‐line flow accretion surveys were also used. The RF model allowed prediction of observed losing and gaining reaches with a high prediction accuracy (“out‐of‐bag” error < 10%). The performance of the model, however, was found to be dependent on the geographic administrative region. The model was also found to be sensitive to slope, distance to headwater, distance to coast, and underlying geological characteristics. When applied in New Zealand, this approach yielded a realistic conceptual representation of national surface–groundwater dynamics that are subsequently being used to inform a national‐scale hydrological model.

This study uses Random Forest algorithm to model students' final year mark in an engineering technology module taught by the University of South Africa. The algorithm uses a supervised learning classification technique to map the different assessment marks and the final mark. Hence, the latter are labelled instances whereas the former constitute the features. Random Forest (RF) has been applied to Structural Analysis 3, which takes into consideration the graduate attribute concept or level of competence as far as assessments are concerned. Firstly, the RF is subjected to imbalanced binary classes, then balanced classes are achieved by Synthetic Minority Oversampling Technique (SMOTE) and class weights adjustment techniques. The results showed that SMOTE brought an improvement in accuracy of 3%. It was also revealed that an increase of 4, 15 and 9% in precision, recall and F1-Score were observed in predicting non-competent students. An increase of 4 and 3% was noticed in the case of the precision and F1-Score respectively in predicting competent students, whereas the recall did not display any change. Despite the RF with SMOTE overperformed standard RF and RF class weights adjustment, all three algorithms were good candidates in the prediction of student performance. RF-SMOTE could be suggested as a guiding instrument when dealing with imbalanced data.

In recent years, various types of terrorist attacks have occurred which have caused worldwide catastrophes. The ability to proactively detect and even predict a potential terrorist risk is critically important for government agencies to react in a timely manner. In this study, a method of geospatial statistics was used to analyse the spatio-temporal evolution of terrorist attacks on the Indochina Peninsula. The machine learning random forest (RF) method was adopted to predict the potential risk of terrorist attacks on the Indochina Peninsula on a spatial scale with 15 driving factors. The RF model performed well with AUC values of 0.839 [95% confidence interval of 0.833–0.844]. The map of the potential distribution of terrorist attack risk was obtained with a 0.05×0.05-degree (approximately 5×5 km) resolution. The results indicate that Thailand is the most dangerous area for terrorist attacks, especially southern Thailand, Bangkok and its surrounding cities. Middle Cambodia and the northern and southern parts of Myanmar are also high-risk areas. Other areas are relatively low risk. This study provides the hotspots for terrorist attacks on a more fine-grained geographical unit. Meanwhile, it shows that machine learning algorithms (e.g., RF) combined with GIS have great potential for simulating the risk of terrorist attacks.