Random Tree Research Articles

An Innovative Path Planning Algorithm for Complex Obstacle Environments with Adaptive Obstacle Density Adjustment: AODA-PF-RRT*

To address the limitations of low node utilization and inadequate adaptability in complex environments encountered by Rapidly-exploring Random Tree (RRT) algorithms during the expansion phase, this study presents an enhanced path planning algorithm—AODA-PF-RRT* (Adaptive Obstacle Density Adjustment-PF-RRT*). The proposed algorithm implements a random extension strategy for nodes that fail collision detection, thereby improving node efficiency. Furthermore, it dynamically partitions the area surrounding sampling points and calculates local obstacle density in real time. By leveraging this density information, the algorithm flexibly adjusts both the number of expansion points and the dichotomy threshold, significantly enhancing its responsiveness to environmental changes. We rigorously demonstrate the algorithm’s probabilistic completeness and asymptotic optimality. Simulation and benchmarking results demonstrate that the AODA-PF-RRT* algorithm not only generates smooth and high-quality paths compared to existing algorithms but also maintains low computational costs in complex environments, showcasing exceptional stability and robustness.

Noisy Dueling Double Deep Q-Network algorithm for autonomous underwater vehicle path planning.

How to improve the success rate of autonomous underwater vehicle (AUV) path planning and reduce travel time as much as possible is a very challenging and crucial problem in the practical applications of AUV in the complex ocean current environment. Traditional reinforcement learning algorithms lack exploration of the environment, and the strategies learned by the agent may not generalize well to other different environments. To address these challenges, we propose a novel AUV path planning algorithm named the Noisy Dueling Double Deep Q-Network (ND3QN) algorithm by modifying the reward function and introducing a noisy network, which generalizes the traditional D3QN algorithm. Compared with the classical algorithm [e.g., Rapidly-exploring Random Trees Star (RRT*), DQN, and D3QN], with simulation experiments conducted in realistic terrain and ocean currents, the proposed ND3QN algorithm demonstrates the outstanding characteristics of a higher success rate of AUV path planning, shorter travel time, and smoother paths.

RRT-guided experience generation for reinforcement learning in autonomous lane keeping

Reinforcement Learning has emerged as a significant component of Machine Learning in the domain of highly automated driving, facilitating various tasks ranging from high-level navigation to control tasks such as trajectory tracking and lane keeping. However, the agent’s action choice during training is often constrained by a balance between exploitation and exploration, which can impede effective learning, especially in environments with sparse rewards. To address this challenge, researchers have explored combining RL with sampling-based exploration methods such as Rapidly-exploring Random Trees to aid in exploration. This paper investigates the effectiveness of classic exploration strategies in RL algorithms, particularly focusing on their ability to cover the state space and provide a quality experience pool for learning agents. The study centers on the lane-keeping problem of a dynamic vehicle model handled by RL, examining a scenario where reward shaping is omitted, leading to sparse rewards. The paper demonstrates how classic exploration techniques often cover only a small portion of the state space, hindering learning. By leveraging RRT to broaden the experience pool, the agent can learn a better policy, as exemplified by the dynamic vehicle model’s lane-following problem.

Trajectory optimization and obstacle avoidance of autonomous robot using Robust and Efficient Rapidly Exploring Random Tree.

One of the key challenges in robotics is the motion planning problem. This paper presents a local trajectory planning and obstacle avoidance strategy based on a novel sampling-based path-finding algorithm designed for autonomous vehicles navigating complex environments. Although sampling-based algorithms have been extensively employed for motion planning, they have notable limitations, such as sluggish convergence rate, significant search time volatility, a vast, dense sample space, and unsmooth search routes. To overcome the limitations, including slow convergence, high computational complexity, and unnecessary search while sampling the whole space, we have proposed the RE-RRT* (Robust and Efficient RRT*) algorithm. This algorithm adapts a new sampling-based path-finding algorithm based on sampling along the displacement from the initial point to the goal point. The sample space is constrained during each stage of the random tree's growth, reducing the number of redundant searches. The RE-RRT* algorithm can converge to a shorter path with fewer iterations. Furthermore, the Choose Parent and Rewire processes are used by RE-RRT* to improve the path in succeeding cycles continuously. Extensive experiments under diverse obstacle settings are performed to validate the effectiveness of the proposed approach. The results demonstrate that the proposed approach outperforms existing methods in terms of computational time, sampling space efficiency, speed, and stability.

Machine Learning Empowered Breast Cancer Diagnosis: Insights from Coimbra Dataset

Aim: The aim of this work is to succinctly communicate the key aspects of a research study on breast cancer. This includes highlighting the global impact and prevalence of breast cancer, emphasizing the challenges of early diagnosis, discussing the potential of technological advancements, and showcasing the role of machine learning algorithms in the context of liver cancer diagnosis. Background: Cancer, notably breast cancer, represents a global health challenge, claiming a significant toll with 12.5% of new cancer cases annually. The prevalence of breast cancer among women worldwide is alarming, resulting in 2.26 million incidents and the unfortunate loss of 685,000 lives. Objective: This article focuses on the critical aspect of early breast cancer diagnosis, acknowledging its heightened difficulty in developing nations compared to developed counterparts. The potential for advancements in technology to serve as a beacon of hope lies in early identification and timely treatment, offering salvation to numerous women and significantly elevating survival chances. Method: In this intricate landscape, machine learning algorithms, particularly in diagnosing liver cancer at its nascent stages, emerge as instrumental tools. The study employs the latest Coimbra dataset, encompassing nine key attributes and a binary classification attribute, with values 1 and 2 signifying benign and malignant cases, respectively. Result: Supervised machine learning algorithms, including Bayes net, multilayer perceptron, IBK, random committee, and random tree, are meticulously applied. Certain models exhibit superior accuracy, precision, recall, and performance, positioning them as promising cornerstones for breast cancer analysis. Conclusion: This structured abstract highlights the urgent need for effective screening and prevention strategies, emphasizing the potential of advanced technology and machine learning algorithms to play a pivotal role in the early detection and analysis of breast cancer, offering hope for improved outcomes and survival rates.

Research on risk detection and management based on machine learning

With the rapid advancement in financial risk management, machine learning techniques are playing an increasingly crucial role in credit assessment and risk detection. They provide financial institutions with more scientific and precise platforms and methodologies for risk management. This paper explores various traditional machine learning methods in risk assessment, introduces the XGBoost ensemble learning method, and integrates traditional machine learning with Long Short-Term Memory (LSTM) neural networks to enhance the generalization capability of risk control and credit assessment models. This approach offers new perspectives and improvement directions for risk assessment standards and practices in the financial industry, affirming the potential application of machine learning technologies in future risk management.The results show that the random forest and decision tree have excellent accuracy and recall rates for distinguishing fraudulent transactions. After the introduction of LSTM neural networks, the accuracy and recall rates of fraudulent transaction recognition models reach around 99%, indicating that these models have good adaptability for fraud risk recognition.

AirNet: predictive machine learning model for air quality forecasting using web interface

Air is one of the most significant elements of the environment. The increasing global air pollution crisis poses an unavoidable threat to human health, environmental sustainability, ecosystems, and the earth's climate. Air pollution has been referred to as a silent killer due to its insidious nature. Its indirect impact on human health further underscores its dangerous effects. Early detection of air quality can potentially save millions of lives globally. A unique and transformative approach can harness the power of machine learning to combat air pollution. This research presents a manual and web-based automatic prediction system that provides real-time alerts on air quality status and can help prevent premature deaths, chronic diseases, and other health problems. Air pollutants, including carbon monoxide (CO), ozone (O3), nitrogen dioxide (NO2), and particulate matter (PM 2.5), are used in this study for feature analysis and extraction. The system utilizes publicly available data from 23,463 different cities worldwide. Data preprocessing was performed before feeding the data into the machine learning models for feature correlation and evaluation. The proposed research uses various machine learning models to predict air quality, including Random Forest (100%), Logistic Regression (79%), Decision Tree (100%), Support Vector Machine (93%), Linear SVC (98%), K-Nearest Neighbor (99%), and Multinomial Naïve Bayes (52%). A user-friendly Django-based web interface offers an accessible platform for users to monitor air quality in real-time, based on the two best-performing models: Random Forest and Decision Tree techniques.

Utility of Certain AI Models in Climate-Induced Disasters

To address the current challenge of climate change at the local and global levels, this article discusses a few important water resources engineering topics, such as estimating the energy dissipation of flowing waters over hilly areas through the provision of regulated stepped channels, predicting the removal of silt deposition in the irrigation canal, and predicting groundwater level. Artificial intelligence (AI) in water resource engineering is now one of the most active study topics. As a result, multiple AI tools such as Random Forest (RF), Random Tree (RT), M5P (M5 model trees), M5Rules, Feed-Forward Neural Networks (FFNNs), Gradient Boosting Machine (GBM), Adaptive Boosting (AdaBoost), and Support Vector Machines kernel-based model (SVM-Pearson VII Universal Kernel, Radial Basis Function) are tested in the present study using various combinations of datasets. However, in various circumstances, including predicting energy dissipation of stepped channels and silt deposition in rivers, AI techniques outperformed the traditional approach in the literature. Out of all the models, the GBM model performed better than other AI tools in both the field of energy dissipation of stepped channels with a coefficient of determination (R2) of 0.998, root mean square error (RMSE) of 0.00182, and mean absolute error (MAE) of 0.0016 and sediment trapping efficiency of vortex tube ejector with an R2 of 0.997, RMSE of 0.769, and MAE of 0.531 during testing. On the other hand, the AI technique could not adequately understand the diversity in groundwater level datasets using field data from various stations. According to the current study, the AI tool works well in some fields of water resource engineering, but it has difficulty in other domains in capturing the diversity of datasets.

Enhancing dietary analysis: Using machine learning for food caloric and health risk assessment.

In the wake of growing concerns regarding diet-related health issues, this study investigates the application of machine learning methods to estimate the energy content and classify the health risks of foods based on the USDA National Nutrient Database. The caloric content of foods was estimated using the nutritional composition (i.e., carbohydrates, protein, total lipid, and total sugar content) and classified based on their weighted health risks, considering their carbohydrate, lipid, and glycemic index levels. The algorithms used for modeling include multiple linear regression (MLR), K-nearest neighbors, support vector machine, random forest regression (RFR), gradient-boosted regression, decision trees (DT), and deep neural networks. The MLR model demonstrated high accuracy on the training dataset (R2=0.99, mean absolute error [MAE]=7.71kcal, and root mean squared error [RMSE]=17.89kcal) and testing dataset (R2=0.99, MAE=7.75kcal, and RMSE=18kcal) in energy estimation, indicating its effectiveness in dietary assessment. The RFR and DT models were useful in categorizing foods into low-health-risk foods, but their performance was reduced in medium and high-health-risk groups. This research contributes to developing tools that could aid in personalized dietary planning and public health interventions to mitigate diet-related health risks. PRACTICAL APPLICATION: This study applies machine learning to estimate how many calories are in food and to understand the health risks different foods might have. By investigating the fats, cholesterol, and sugars in food items listed in a public database, we can better plan diets or develop apps that help people make healthier eating choices. This work aims to improve public access to nutritional information, supporting efforts to combat diet-related diseases through educational materials and applications that guide dietary choices in various settings.

Enhanced MRI brain tumor detection and classification via topological data analysis and low-rank tensor decomposition

The advent of artificial intelligence in medical imaging has paved the way for significant advancements in the diagnosis of brain tumors. This study presents a novel ensemble approach that uses magnetic resonance imaging (MRI) to identify and categorize common brain cancers, such as pituitary, meningioma, and glioma. The proposed workflow is composed of a two-fold approach: firstly, it employs non-trivial image enhancement techniques in data preprocessing, low-rank Tucker decomposition for dimensionality reduction, and machine learning (ML) classifiers to detect and predict the type of brain tumor. Secondly, persistent homology (PH), a topological data analysis (TDA) technique, is exploited to extract potential critical areas in MRI scans. When paired with the ML classifier output, this additional information can help domain experts to identify areas of interest that might contain tumor signatures, improving the interpretability of ML predictions. When compared to automated diagnoses, this transparency adds another level of confidence and is essential for clinical acceptance. The performance of the system was quantitatively evaluated on a well-known MRI dataset, with an overall classification accuracy of 97.28% using an extremely randomized trees model. The promising results show that the integration of TDA, ML, and low-rank approximation methods is a successful approach for brain tumor identification and categorization, providing a solid foundation for further study and clinical application.

Predicting Livestock Farmers’ Attitudes towards Improved Sheep Breeds in Ahar City through Data Mining Methods

Sheep breeding is one of the most important economic activities in Ahar City, Iran. However, due to traditional production techniques, livestock farmers face the problem of low productivity. To address this issue, traditional breeds can be replaced with improved and high-yielding ones; in the first stage, this requires the acceptance of these new sheep breeds by the region’s ranchers. This research aimed to evaluate the attitudes of the livestock breeders of Ahar City towards the improved breeds of sheep and the influential factors. We collected data through in-person interviews using a simple random sampling method, surveying 100 sheep breeders in Ahar. The breeders were categorized into three groups based on their attitudes towards improved breeds: negative, indifferent, and positive. Next, we employed data mining-based methods, including multilayer perceptron neural networks, random forest, and random tree algorithms. These helped identify essential variables affecting ranchers’ attitudes. The results showed that several factors contribute to the ranchers’ philosophy, with the number of sheep sold in the past year and the total sheep ownership being the most significant ones. Comparing statistical evaluation criteria, we found that the random tree algorithm outperformed other methods in predicting and classifying livestock farmers, achieving a prediction accuracy rate of 86% for a sample of 100 farmers. Based on our findings, promoting training courses and raising awareness about the benefits of breeding new sheep breeds, along with providing facilities and credits based on economic conditions, can foster a positive attitude among herders. Increasing the number of sheep owned and improving marketing strategies can further enhance this positive outlook.

Incorporating Crumb Rubber in Slag-Based Geopolymer: Experimental Work and Predictive Modelling

Crumb rubber (CR), a prevalent hazardous waste material, poses significant environmental challenges that necessitate innovative management strategies. Recent studies have focused on incorporating CR and ground granulated blast furnace slag (GGBS) into geopolymer concrete (GC) to address this challenge. This research presents a novel approach that combines experimental analysis with machine learning (ML) techniques to investigate the chemical effects of these materials on GC. Initially, compressive strength (CS) tests are conducted on rubberized geopolymer (RG) concrete, and the resulting data are further analyzed through ML algorithms to enhance the understanding of material behavior. Despite numerous attempts by researchers to integrate CR into geopolymer-based materials, the challenge of mitigating strength degradation persists. Therefore, it becomes imperative to construct a predictive model that relies on new variables influencing the strength characteristics. This research develops a predictive model to assess concrete compressive strength (CS). Key variables such as CR grade (particle size), incorporation percentage, and oxide ratio were analyzed for their impact on geopolymer strength. Employing ensemble techniques, namely Bagging (BA) and Gradient boosting (GB) with five learners (M5P, random forest (RF), random tree (RT), reduced error pruning tree (REPT), and support vector machine (SVM), the study found that M5P-GB models offered the most accurate CS predictions with the highest accuracy and lowest errors. Moreover, the dataset was checked via the k-fold cross-validation technique and confirmed the developed models' robustness, reliability, and effectiveness. Furthermore, uncertainty analysis revealed that GB-M5P-unpruned models-maintained uncertainty percentages of 14.769% at 7 days and 11.277% at 28 days, which was under the 35% threshold. Additionally, the sensitivity analysis identified the CR grade and replacement percentage by volume of crusher dust (CD) as the most critical factors affecting CS. These findings underscore the importance of considering predictive model development in decision-making processes related to RG concrete properties, providing valuable insights into optimizing the model's robustness and reliability for practical applications. However, the study's reliance on specific input parameters and controlled laboratory conditions may limit the generalizability of the predictive models, necessitating further validation under diverse real-world conditions and variations in CR source and environmental factors.

KF-PLS: Optimizing Kernel Partial Least-Squares (K-PLS) with Kernel Flows

Partial Least-Squares (PLS) regression is a widely used tool in chemometrics for performing multivariate regression. As PLS has a limited capacity of modelling non-linear relations between the predictor variables and the response, Kernel PLS (K-PLS) has been introduced for modelling non-linear predictor-response relations. Most available studies use fixed kernel parameters, reducing the performance potential of the method. Only a few studies have been conducted on optimizing the kernel parameters for K-PLS. In this article, we propose a methodology for the kernel function optimization based on Kernel Flows (KF), a technique developed for Gaussian Process Regression (GPR). The results are illustrated with four case studies. The case studies represent both numerical examples and real data used in classification and regression tasks. K-PLS optimized with KF, called KF-PLS in this study, is shown to yield good results in all illustrated scenarios, outperforming literature results and other non-linear regression methodologies. In the present study, KF-PLS has been compared to convolutional neural networks (CNN), random trees, ensemble methods, support vector machines (SVM), and GPR, and it has proved to perform very well.

Machine learning-assisted SCAPS device simulation for photovoltaic parameters prediction of [formula omitted] perovskite solar cells

Perovskite solar cells (PSCs) have garnered significant research attention because of their remarkable surge in power conversion efficiency (PCE) surpassing 26% within a short timeframe. However, the widespread adoption of these highly efficient PSCs is hampered by the inherent toxicity associated with lead (Pb)-based perovskites, which currently dominate the field. Researchers have explored alternative cations like Ge2+ and Sn2+ Owing to their identical oxidation states to Pb2+, paving the way for lead-free PSC development. By examining the critical factors influencing CsSnI3 thin film solar cells’ performance as well as their interdependencies, this work seeks to improve the efficiency of these devices. To do so, machine learning algorithms (ML) are employed to figure out the critical factors influencing the completion of CsSnI3 solar cells. A new dataset of 7870 data points is generated through simulations with SCAPS-1D to effectively depict the perovskite solar cell. Five ML regression algorithms are thereafter utilized to predict the photovoltaic parameters of solar cells, such as the fill factor (FF), open circuit voltage (VOC), short circuit current (JSC), and PCE. The study shows that random forest (RF) and decision tree (DT) are the best performing algorithms, among which DT produces high accuracy and low error for predicting the PCE with a coefficient of determination R2 of 0.99, Pearson’s coefficient (Rvalue) of 0.99, Spearman’s correlation coefficient (ρ) of 0.99 and root mean square error of 0.08 on the testing set. Finally, this study offers valuable guidance for further experiment optimization by shedding light on the optimal device dimensions and essential components.

Leveraging machine learning to minimize experimental trials and predict hot deformation behaviour in dual phase high entropy alloys

In recent time, high entropy alloys (HEAs) are widely used due to their wide design space and remarkable properties allowing a vast range of property variations with myriads of possible combinations of constituent elements. This research utilizes machine learning techniques, including Random Forest (RF), K-Nearest Neighbors (KNN), XGBoost (XGB), Decision Tree (DT), and Support Vector Regressor (SVR), to predict the flow stress behavior of the dual FCC phase CoCrCu1.2FeNi high entropy alloy (HEA) at new temperature and strain rates to reduce the experiments dependency. The RF and KNN models were identified as top performers, with R² values of 0.987 and 0.986, respectively. These models successfully predicted flow stresses, aligning closely with actual measurements. A new flow stress-strain curve was generated at new temperature of 1123K, demonstrating good performance metrics for RF and KNN. with R² values of 0.97 and 0.958 respectively. This approach has the potential to significantly reduce the need for extensive experimental work in determining the mechanical behaviour of HEA, offering substantial savings in time, cost, and energy, and marks a significant advancement in the development of such alloys.

Identifying Crop Distress and Stress-Induced Plant Diseases Using Hyper Spectral Image Analysis

The purpose of this study is to determine whether hyper spectral images can be used to identify plant diseases and crop pressure from aerial photographs. With extensive research on the prevalent methods used closer to the problem, this study offers a potent strategy for identifying crop distress and illnesses using this efficient imaging technique. To identify the spectral fingerprints of common indications and symptoms of plant diseases and crop strains, this study evaluates the available hyper spectral photo datasets. After that, the data are examined using two learning algorithms—the highly randomized trees and the Random Woodland set of rules—to create predictions that are entirely dependent on the results that are discovered. In the end, a benchmarked set of test statistics is used to assess the prediction accuracy. The results of this study show that hyper spectral photo evaluation has a strong and promising utility for crop stress and disease identification. Hyper spectral light evaluation is a method for identifying plant diseases caused by strain on crops. By gathering and evaluating high-dimensional spectral reflection data from satellite or aircraft structures, details regarding the physiological homes of flowers can be identified. The health of plants, illnesses brought on by stress, and agricultural productivity predictions can all be made using these facts. Hyper spectral recordings can also be used to create actions that reduce agricultural losses and enhance the health of plants that are prone to disease.

Estimating high-resolution snow depth over the North Hemisphere mountains utilizing active microwave backscatter and machine learning

While ground meteorological stations provide accurate snow depth data, their limited spatial coverage results in observational gaps. Satellites offer long-term, large-scale observations, addressing these gaps. Existing snow depth retrieval algorithms mainly use passive microwave remote sensing data with a 25 km resolution, insufficient for capturing snow depth variability in mountainous areas. This paper introduces active microwave backscatter data and machine learning techniques for high-resolution snow depth estimation. We conducted a preliminary exploration of the relationship between Sentinel-1 backscatter coefficient σ0 and snow depth. Due to factors such as vegetation coverage and underlying soil properties, the relationship between σ0 and snow depth is complex and nonlinear. Consequently, six machine learning models were trained to learn this relationship using σ0 and auxiliary data as input features, with in-situ snow depth serving as the target variable. After extensive validation, the Extreme Random Trees (ERT) model was selected for its high accuracy and stability. Using the ERT model, we generated 500 m-resolution snow depth data for northern hemisphere mountains, then analyzed temporal snow depth variations and altitudinal stratification.

Comparison of hybrid deep learning models for estimation of the time-dependent scour depth downstream of river training structures

Submerged weirs, mainly positioned downstream of bridges, play a key role in safeguarding against floods and long-term scour damage. However, the structural stability of these structures could be threatened by local scour holes. This study evaluates five deep learning algorithms—Deep Neural Networks, Convolutional Neural Networks (CNNs), Convolutional Extreme Gradient Boosting (CXGB), convolutional extremely randomized trees regression, and Self-Attention-based Convolutional Neural Network (SA-CNN) in predicting the evolution of scour depth. Using Hyperband and Bayesian optimization, the models were fine-tuned for maximum accuracy. Additionally, this study investigates the impact of two data splitting methods, including random pointwise sampling and case-wise sampling on model performance. Results indicate that the hybrid CXGB and the SA-CNN models outperform other models in terms of accuracy of the estimation of the time-dependent scour depth with R2 = 0.997 in pointwise and R2 = 0.878 in case-wise split strategies, respectively. This not only demonstrates the effectiveness of these sophisticated algorithms in time-dependent scour estimation but also clarifies the effects of various data sampling techniques on model performance. Finally, the contribution of features in provided estimations is discussed utilizing SHapley Additive exPlanations values. Results indicated that the time (T) and the ratio of the flow velocity to critical velocity U0/Uc had the greatest effect on the model outputs, while side slopes indicated a negligible effect on model output compatible with the physics of the problem.

Machine Learning approach in the prediction of Fog: An Early Warning System.

The aviation sector is extremely vulnerable to fog. Thus, accurate fog predictions are essential foraviation sector efficiency, particularly airport management and flight scheduling. Even with numerical weather predictionmodels and guiding systems, fog prediction is challenging. The difficulty of fog prediction is due to the inability to graspthe micro-scale factors that cause fog to form, intensify, augment and dissipate in the boundary layer. This study looks athow well machine learning (ML) tools can predict fog (Visibility <1000 m) and dense fog (Visibility <200 m) at India'sJay Prakash Narayan International Airport (ICAO Index-VEPT), a representative station of the Indo-Gangetic Plains(IGP). The proposed ensemble ML-based model was trained using hourly synoptic data from 2014 to 2020 and testedusing data from 2021 to 2022 (December to February). Once the features are chosen and the forecasters' local knowledgeis taken into account, the dry bulb temperature (°C), dew point temperature (°C), relative humidity (%), cloud amount(octa), wind direction (degrees from true north) and wind speed (knots) are used to build the proposed ML models. MLalgorithms were trained on meteorological data from 1500 to 2200 UTC to predict fog (Visibility <1000 m) and densefog (Visibility <200 m) for the next day at 0000 UTC, with a two-hour lead time. For fog forecasting at 0000 UTC with a 4-hour lead time, ML models were trained with data from 1300 to 2000 UTC and so on. This study evaluates parameter tuning in six level 0 ML models: distributed random forest (DFR), deep learning (DL), gradient boosting machine (GBM), generalized linear model (GLM), extremely randomized tree (XRT), XG Boost and stacked ensemble at level 1. The performance metrics and statistical skill scores indicate that DRF and DL models perform well for lead times of 2 and 4 hours at level 0 for fog (visibility <1000 m) and dense fog (visibility <200 m). But the proposed ensemble models outperform all the base models at level 0 and are recognized as the best instrument for predicting fog at Patna Airport.

Investigation of wettability and IFT alteration during hydrogen storage using machine learning

Reducing the environmental impact caused by the production or use of carbon dioxide (CO2) and other greenhouse gases (GHG) has recently attracted the attention of scientific, research, and industrial communities. In this context, oil production and enhanced oil recovery (EOR) have also focused on using environmentally friendly methods. CO2 has been studied as a significant gas in reducing harmful environmental effects and preventing its release into the atmosphere. This gas, along with methane (CH4) and nitrogen (N2), is recognized as a ‘cushion gas’. Given that hydrogen (H2) is considered a green and environmentally friendly gas, its storage for altering wettability (contact angle (CA) and interfacial tension (IFT)) has recently become an intriguing topic. This study examines how H2 can be utilized as a novel cushion gas in EOR systems. In this research, the role of H2 and its storage in altering wettability in the presence of other cushion gases has been investigated. The performance of H2 in changing the CA and IFT with other gases has also been compared using machine learning (ML) models. During this process, ML and experimental data were used to predict and report the values of IFT and CA. The data used underwent statistical and quantitative preprocessing, processing, evaluation, and validation, with outliers and skewed data removed. Subsequently, ML models such as Random Forest (RF), Random Tree, and LSBoost were implemented on training and testing data. During this process of modeling and predicting IFT and CA, the hyperparameters were optimized using Bayesian algorithms and random search (RS) methods. Finally, the results and performance of the modeling were evaluated, with the LSBoost modeling method using Bayesian optimization reporting R2 values of 0.998614 for IFT and 0.986999 for CA.