Regression Tree Method Research Articles

Economic and environmental factors determining spatial variation in soil liming in Poland

The aim of this research was to identify the key factors determining the level of soil liming in Poland. Spatial analyses were conducted using data from the 2020 National Agricultural Census. The study presents regional variations in selected parameters that characterise agricultural production quality and structural features, including the market connections of farms. The Classification and Regression Trees (C&RT) method was employed to identify the factors that most significantly influence the intensity of liming in the examined districts. These factors include the share of permanent grasslands, labour inputs, the percentage of farms engaged in commercial production, farm size, and the level of mineral fertilisation (NPK). The results indicate that regions with larger farms and higher NPK fertilisation tend to apply more lime. In contrast, areas with a higher share of permanent grasslands and labour-intensive activities show lower levels of liming. These findings suggest that improving agricultural practices, particularly in regions with small farms and limited fertilisation, could enhance soil liming efforts and align them with the European Green Deal's goals for sustainable agriculture.

Investigation of Higher Education Students' Interest in Sports and Usage Levels of University Sports Facilities

Introduction The size, number, capacity, characteristics and quality of the areas where sports activities can be carried out in universities are among the leading indicators of student investment. Objective: To determine the interest and attitudes of university students in the Western Black Sea region towards sports, their level of utilisation of university sports facilities, and whether university sports facilities significantly predict students' interest and attitudes towards sports. Methodology: The classification and regression tree method (CART), one of the data mining methods, was used through a relational screening model. Data were collected from 1462 associate and undergraduate students (789 males and 673 females) studying in different departments of 7 universities in the Western Black Sea Region in 2022-2023 with two scales of Ceyhun (2015, 2017). Results: It was determined that the areas where university students do sports, their level of interest in sports, their age, class level, duration of doing sports and their active participation in sports affect their level of utilisation of sports facilities. Discussions: It was observed that the level of interest in sports of female university students was higher than that of males, and the level of interest in sports of students who do sports was negatively correlated with the duration of doing sports. Conclusion: The efficient use of sports facilities, one of the main factors of sports participation in universities, is directly proportional to the university's facilities and individual factors.

Regression Trees With Fused Leaves.

We propose a novel regression tree method named "TreeFuL," an abbreviation for 'Tree with Fused Leaves.' TreeFuL innovatively combines recursive partitioning with fused regularization, offering a distinct approach to the conventional pruning method. One of TreeFuL's noteworthy advantages is its capacity for cross-validated amalgamation of non-neighboring terminal nodes. This is facilitated by a leaf coloring scheme that supports tree shearing and node amalgamation. As a result, TreeFuL facilitates the development of more parsimonious tree models without compromising predictive accuracy. The refined model offers enhanced interpretability, making it particularly well-suited for biomedical applications of decision trees, such as disease diagnosis and prognosis. We demonstrate the practical advantages of our proposed method through simulation studies and an analysis of data collected in an obesity study.

Analyzing the rainfed wheat yield gap in Northwest Iran

Wheat is one of the major crops in Iran, covering more than 36% of the total croplands and more than 70% of rainfed farmlands. However, the trend of crop yield in the past decades does not show promising performance and the average yield of rainfed wheat in Iran is as low as 1250 ​kg ​ha−1. Therefore, quantifying the rainfed wheat yield gap, and determining its causes and importance could help wheat self-sufficiency in Iran. Boundary line analysis (BLA) and regression tree (RT) models were employed to analyze the yield gap and find out the influencing factors. For this purpose, an on-farm survey was conducted in 210 rainfed winter wheat fields in East Azerbaijan province during the 2020–21 cropping season. Data were collected from local experts and farmers concerning crop yield and related management practices. Actual farmers’ yields were 1095 ​kg ​ha−1. BLA showed a significant yield gap ranging from 1755 to 2186 ​kg ​ha−1 (61.5–66.6% of attainable yield) and determined the attainable yield of 3072 ​kg ​ha−1. The BLA indicated that, among the quantitative independent variables, nitrogen fertilizer and planting density management were the important managerial practices responsible for the yield gaps. Additionally, the pronounced differences in crop yield across various rotation systems and varieties highlight their critical role in realizing crop attainable yield as the qualitative independent variables. By the regression tree (RT) model, the yield gap was determined about 65%. Crop rotation, crop variety, and sowing date were determined as key factors influencing the rainfed wheat yield gap. Findings showed that the combination of BLA and RT methods can be used effectively to quantify the crop yield gap and its influencing factors. Regarding the large yield gap, there is a good opportunity to obtain higher yields by optimizing managerial practices/inputs.

Unravelling the knowledge matrix: exploring knowledge-sharing behaviours on market-based platforms using regression tree analysis

PurposeThis study aims to enhance the effectiveness of knowledge markets and overall knowledge management (KM) practices within organisations. By addressing the challenge of internal knowledge stickiness, it seeks to demonstrate how machine learning and AI approaches, specifically a text-based AI method for personality assessment and regression trees for behavioural analysis, can automate and personalise knowledge market incentivisation mechanisms.Design/methodology/approachThe research employs a novel approach by integrating machine learning methodologies to overcome the limitations of traditional statistical methods. A natural language processing (NLP)-based AI tool is used to assess employees’ personalities, and regression tree analysis is applied to predict and categorise behavioural patterns in knowledge-sharing contexts. This approach is designed to capture the complex interplay between individual personality traits and environmental factors, which traditional methods often fail to adequately address.FindingsCognitive style was confirmed as a key predictor of knowledge-sharing, with extrinsic motivators outweighing intrinsic ones in market-based platforms. These findings underscore the significance of diverse combinations of environmental and individual factors in promoting knowledge sharing, offering key insights that can inform the automatic design of personalised interventions for community managers of such platforms.Originality/valueThis research stands out as it is the first to empirically explore the interaction between the individual and the environment in shaping actual knowledge-sharing behaviours, using advanced methodologies. The increased automation in the process extends the practical contribution of this study, enabling a more efficient, automated assessment process, and thus making critical theoretical and practical advancements in understanding and enhancing knowledge-sharing behaviours.

Development of a recommendation engine for university study programme selection: A regression-based approach

This paper is the third in a series focused on bridging the gap between secondary and higher education. Our primary objective is to develop a robust theoretical framework for an innovative e-business model called the Undergraduate Study Programme Search System (USPSS). This system considers multiple criteria to reduce the likelihood of exam failure or the need for multiple retakes, while maximizing the chances of successful program completion. Testing of the proposed algorithm demonstrated that the Stochastic Gradient Boosted Regression Trees method outperforms the current method used in Lithuania for admitting applicants to 47 educational programs. Specifically, it is more accurate than the Probabilistic Neural Network for 25 programs, the Ensemble of Regression Trees for 24 programs, the Single Regression Tree for 18 programs, the Random Forest Regression for 16 programs, the Bayesian Additive Regression Trees for 13 programs, and the Regression by Discretization for 10 programs.

Short-term Load Forecasting using Genetic Algorithm based Artificial Neural Network

The electrical load Forecasting is considered one of the main and important points for the planning for the future in the electric power networks in both of transmission and distribution systems. In this paper, the genetic optimization algorithm (GA) with the artificial intelligent network (ANN) methods are proposed to forecast the electrical load of the building of Electrical Power Distribution office in the Iraq /Anbar Governorate. Moreover, the optimization method of GA is presented to train the ANN weights using Matlab software with simple programming code. Then, the proposed method is applied to the actual load and the results are compared with that obtained in traditional method such as ANN and regression trees (RG trees) methods. Hence, from the simulation results and the comparison, we can confirm that the proposed method is better than the traditional methods, and it achieved a good agreement with actual load and it provide a good total RMSE of 6% when compared with other methods

Examining the relationship between the built environment and carbon emissions from operating vehicles: enlightenment from nonlinear models.

Carbon emissions from urban transportation significantly contribute to overall transportation emissions and are a major cause of the continuous rise in global temperatures. Understanding the spatial distribution and influencing factors of carbon emissions from operating vehicles can aid in formulating targeted policies and promoting emission reduction. To analyze the factors influencing urban traffic carbon emissions, we calculated emissions using trajectory data from operating vehicles in Shenzhen. We then used gradient boosting regression tree methods, specifically RF, XGBoost, and LightGBM models, to analyze the impact of the built environment on vehicle emissions. We used the XGBoost model for detailed factor analysis by comparing the models. The results indicate that bus stops, intersections, housing density, metro stops, and land use mix are the top five factors influencing emissions. When road density is 0-15 km/km2, the distance from the city center is 0-6 km, and the population exceeds 2000/km2, the built environment significantly reduces vehicle emissions.

Classification and regression tree (CART) for predicting cadmium (Cd) uptake by rice (Oryza sativa L.) and its application to derive soil Cd threshold based on field data

The entry of Cd into soil-rice systems is a growing concern as it can pose potential risks to public health. To derive regional soil Cd threshold, a total of 333 paired soil and rice samples was collected in Anhui Province, Eastern China. The results showed that the total soil Cd and soil Zn/Cd were the most significant variables contributing to Cd content in polished rice. The Chinese Soil Quality Standards might overestimate risk posed by Cd-contaminated soil for rice production in the mining area due to high Zn/Cd values of some mining-associated soils. Cd levels in polished rice can be predictable using stepwise multiple linear regression (MLR) model. However, the derived soil Cd threshold based on the MLR model would be unrealistically high. The classification and regression tree method (CART) performed well in simulating Cd levels in polished rice and can be used to derive soil Cd threshold instead of MLR to minimize the uncertainty.

Disentangling the Impacts of Environmental Factors on Evaporative Fraction Across Climate Regimes

AbstractEvaporative fraction (EF) is a useful measure for quantifying land surface energy partitioning processes and determining evaporative regimes; however, its influencing factors remain highly uncertain. Here, global data sets were compiled to disentangle the effects of environmental variables on EF variations along climate and land surface gradients. We found that (a) at annual timescales, ecosystem‐level EF could be expressed as a power law function of aridity index. The relationships of mean annual soil water content (SWC) and leaf area index (LAI) with mean annual EF resembled the traditional evaporative regime theory; (b) at daily timescales, the boosted regression tree method quantitatively revealed that the impacts of environmental variables (including meteorological variables) on EF showed equal importance, especially at humid sites, primarily due to the different response direction and magnitude of latent heat (LE) and sensible heat (H) fluxes to environmental changes. Particularly, the contrasting responses of LE (positive) and H (negative) to SWC, LAI, and relative humidity enhanced the positive effects of those influencing variables on EF; whereas, the correlations between EF and energy‐related factors (i.e., net radiation‐Rn and air temperature‐Ta) deteriorated as both LE and H showed positive response patterns to those variables; (c) meteorological factors were also found to have nonlinear effects on daily EF, further modified by climatic conditions. Rn near 150 W/m2 and Ta near 15°C appeared to be important energy‐partitioning thresholds at drier and humid sites, respectively. Moreover, changing interactions among environmental variables with climates were demonstrated to be important for better explaining EF variations.

Evaluating missing data handling methods for developing building energy benchmarking models

This study explored methods for handling missing data in the development of machine learning-based energy benchmarking models, assessing their training time, performance, and variance. Unlike the common assumption of missing completely at random, this study adopted a missing at random (MAR) perspective, which is more appropriate for building data. We compared the inherent missing data handling method of extreme gradient boosting (XGBoost) with the Median, k-nearest neighbors (KNN), and classification and regression trees (CART) methods, alongside Shapley additive explanation (SHAP) method for model interpretability. The findings indicate that, despite its computational demands, the CART method most accurately mirrors the original data distribution, thereby enhancing model performance and stability. The KNN method is effective, while the XGBoost method is viable under computational time constraints. This work highlights the importance of reliable test data for performing accurate evaluations of imputation methods. These results offer guidelines for the selection of imputation methods in model development, contributing to the improved accuracy of energy benchmarking models. The MAR-based approach for missing data analysis holds promise for future research on building energy data, providing crucial insights for accurate energy benchmark model performance assessments.

Integration of prediction and optimization for smart stock portfolio selection

Machine learning (ML) algorithms pose significant challenges in predicting unknown parameters for optimization models in decision-making scenarios. Conventionally, prediction models are optimized independently in decision-making processes, whereas ML algorithms primarily focus on minimizing prediction errors, neglecting the role of decision-making in downstream optimization tasks. The pursuit of high prediction accuracy may not always align with the goal of reducing decision errors. The idea of reducing decision errors has been proposed to address this limitation. This paper introduces an optimization process that integrates predictive regression models within a mean–variance optimization setting. This innovative technique introduces a general loss function to capture decision errors. Consequently, the predictive model not only focuses on forecasting unknown optimization parameters but also emphasizes the predicted values that minimize decision errors. This approach prioritizes decision accuracy over the potential accuracy of unknown parameter prediction. In contrast to traditional ML approaches that minimize standard loss functions such as mean squared error, our proposed model seeks to minimize the objective value derived directly from the decision-making problem. Furthermore, this strategy is validated by developing an optimization-based regression tree model for predicting stock returns and reducing decision errors. Empirical evaluations of our framework reveal its superiority over conventional regression tree methods, demonstrating enhanced decision quality. The computational experiments are conducted on a stock market dataset to compare the effectiveness of the proposed framework with the conventional regression tree-based approach. Remarkably, the results confirm the strengths inherent in this holistic approach.

Spatio‐temporal analysis of snow depth and snow water equivalent in a mountainous catchment: Insights from in‐situ observations and statistical modelling

AbstractThis research, conducted in the mountainous catchment near Abant Lake in the Western Black Sea region of Türkiye, aimed to investigate the spatiotemporal variations of snow depth (SD) and snow water equivalent (SWE) throughout the snow season from December 2019 to March 2020, encompassing both accumulation and melting periods. In total, 14 snow surveys were conducted, covering 58 permanent snow measurement points (PSMP) marked with snow poles. The classification and regression tree (CART) method was employed to statistically analyse their relationships with eight variables: snow period, forest canopy, aspect, slope, elevation, slope position, plan and profile curvature. The root mean square error (RMSE) for SD and SWE was determined to be 0.15 m and 46 mm, respectively. The study findings revealed that mean SD and SWE values were higher in forest gaps compared with under‐forest and open areas. Although the snow cover disappeared earliest in under‐forest areas, the melting rate was observed to be 43% and 17% slower compared with forest gaps and open areas, respectively. Wind redistribution resulted in minimum snow accumulation on western aspects, upper slope positions and ridges, while maximum accumulation was observed on southern aspects, valleys and lower slope positions. Higher elevations (>1580 meters) experienced faster snow melting rates, leading to earlier disappearance of snow cover. PSMPs located on slopes with lower degrees (<15°) exhibited lesser accumulation and earlier snow disappearance. The CART model identified the snow period as the most significant factor in predicting SD and SWE, based on variations in snowfall and air temperature. Other significant variables included forest canopy, aspect and elevation. The study suggests that the CART method is well‐suited for modelling complex snow dynamics, providing valuable insights into spatiotemporal variations in SD and SWE in mountainous regions.

The predictive role of lung clearance index on FEV1 decline in cystic fibrosis.

The lung clearance index (LCI) is a sensitive lung function index that is used to detect early lung disease changes in children with cystic fibrosis (CF). This study aimed to define the predictive role of baseline LCI, along with other potential factors on the change in forced expiratory volume in one second (FEV1) during one-year follow-up in CF patients who had a percent predicted (pp) FEV1≥80. LCI was concurrently performed on 57 CF patients who had ppFEV1 ≥80 at month zero. The ppFEV1 decline was evaluated prospectively during the one year follow up. The primary outcome of ppFEV1 decline in the study group in one year was dichotomized according to the median value for the decline in ppFEV1, which was 3.7. The LCI value predicting ppFEV1 decline at the end of one year was calculated with receiver operating characteristic curve analysis. Regression analysis was performed. Furthermore, a decision tree was constructed using classification and regression tree methods to better define the potential effect of confounders on the ppFEV1 decline. The LCI value for predicting ppFEV1 decline >3.7% at the end of one year was 8.2 (area under the curve: 0.80) Multivariable regression analysis showed that the absence of the F508del mutation in at least one allele, LCI >8.2 and initial FEV1 z-score were predictors of a ppFEV1 decline >3.7 (p<0.001). Factors altering ppFEV1 decline>3.7% at the end of one-year evaluated by decision trees were as follows: initial FEV1 z-score, type of CFTR mutation, LCI value and initial weight-for-age z-score. LCI is sensitive for predicting ppFEV1 decline in patients with ppFEV1 ≥80 along with the initial FEV1-z-score and type of CFTR mutation.

Enhancing TikTok Account Performance with Data Pattern Identification

TikTok is a social media platform widely used to share information through short videos to achieve goals and interests in business, education, politics, government, and personal existence. Every activity on this platform is recorded and presented privately to each account owner. However, this data has not been utilized optimally to improve account performance. This research aims to offer a data analysis concept that integrates statistical and machine learning approaches to identify data patterns in each user's data collection, enabling the improvement of account performance. The approach utilizes Linear Regression, k-means, and Decision Tree methods. The results obtained show that the concept of identifying data patterns in TikTok account data has successfully developed a predictive model for video posts that can potentially increase total viewership, video plays, and audience engagement. This is achieved through optimizing video components such as captions, text, hashtags, sound genre, and video type. The outcome yielded a classification model that can predict capable component content to enhance account performance.

Predicting Student Achievement via Machine Learning: Evidence from Turkish Subset of PISA

This study seeks to identify the determinants of academic performance in mathematics, science, and reading among Turkish secondary school students. Using data from the OECD's PISA 2018 survey, which includes several student- and school-level variables as well as test scores, we employed a range of supervised machine learning methods specifically ensemble decision trees to assess their predictive performance. Our results indicate that the boosted regression tree (BRT) method outperforms other methods bagging and random forest regression trees. Notably, the BRT highlights the importance of general secondary education programs over vocational and technical (VAT) education in predicting academic achievement. Moreover, both characteristics specific to student and school environment are demonstrated to be significant predictors of academic performance in all subject areas. These findings contribute to the development of evidence-based educational policies in Turkey.

Sociodemographic and psychological factors affecting motor vehicle crashes (MVCs): a classification analysis based on the contextual-mediated model of traffic-accident involvement

The study aimed to determine the sociodemographic and psychological profiles of drivers with a history of motor vehicle crashes (MVCs), following the contextual-mediated model of crash involvement, and trying to define similarities and differences with drivers without MVCs. Although road trauma prevention has become a central public health issue, the study of psychological determinants of MVCs does not have consistent results due to methodological and theoretical weaknesses. Three-hundred and forty-five active drivers (20% females) completed an extensive office-based fitness-to-drive evaluation including measures of cognition, personality, self-reported driving-related behaviors, attitudes, as well as computerized measures of driving performance. The Classification and Regression Tree method (CART) was used to identify discriminant predictors. The classification identified several relevant predictors; the personality trait of Discostraint (as a distal context variable; cut-point: 50 T points) and motor speed (as a proximal context variable; cut-point: 64 percentile ranks). The global classification model increased approximately 3 times the probability of identifying people with a history of MVC involvement, starting from an estimated prevalence of being involved in an MVC in a period of five years in the population of active drivers. Consistent with the ‘contextual-mediated model of traffic accident involvement’, the results of the present study suggest that road trauma analysis should focus on both distal and proximal driver-related factors by paying attention to their association in determining MVCs. These results represent a valuable source of knowledge for researchers and practitioners for preventing road trauma.

An integrated data envelopment analysis and regression tree method for new product price estimation

An integrated data envelopment analysis and regression tree method for new product price estimation

Classification of Land Suitability For Soybean Crops Using The Cart Method and Feature Selection Using an Algorithm ABC

The allocated area for soybean cultivation has been gradually decreasing, leading to a decline in both production and productivity. Consequently, the current level of soybean production and productivity falls short of meeting the demand within the community. One potential solution to augment soybean output and efficiency involves allocating specific parcels of land for soybean cultivation. It is essential to conduct land evaluations tailored to soybean cultivation, accounting for the land's inherent potential, in order to optimize land utilization. Thus, a comprehensive system is required to assess land suitability, particularly for soybean cultivation, and employ the results of this classification as recommendations for land allocation. This research employess combination the Classification and Regression Tree (CART) method and the Artificial Bee Colony (ABC) algorithm to classify suitable land for soybean cultivation. CART is used for classification and ABC is utilized for feature selection to identify the most relevant attributes in case of the algorithm improvement. Through a series of iterative experiments involving 5, 10, 25, 50, 75, and 100 iterations, the best attribute was determined following three attempts at each iteration. The Confusion Matrix test yielded an accuracy rate of 94.22% for the CART method in the second experiment, while the combined use of the best ABC and CART combination resulted in an accuracy rate of 97.11%. Therefore, it can be concluded that the integration of the artificial bee colony (ABC) algorithm with the classification and regression tree (CART) method outperforms the sole use of the CART method in terms of accuracy.

Machine learning‐based streamflow forecasting using CMIP6 scenarios: Assessing performance and improving hydrological projections and climate change

AbstractWater is essential for humans as well as for all living organisms to sustain their lives. Therefore, any climate‐driven change in available resources has significant impacts on the environment and life. Global climate models (GCMs) are one of the most practical methods to evaluate climate change. Based on this, this research evaluated the capability of GCMs from the Coupled Model Intercomparison Project 6 (CMIP6) to reproduce the historical flow of climate prediction centre data for the Konya Closed basin and to project the climate of the basin using the selected GCMs. Global climate models based on the CMIP6 under the scenario of common socioeconomic pathways (SSP245 and SSP 585) were used to analyse the climate change effect on streamflow of the study area by Bias Correction of GCM Models using Long Short‐Term Memory (LSTM), Bidirectional LSTM (BiLSTM), AdaBoost, Gradient Boosting, Regression Tree, and Random Forest methods. The coefficient of determination (R2), mean square error (MSE), mean absolute error (MAE), root mean square error (RMSE) were used to assess the performance of the methods. Findings show that the Random Forest Model consistently outperformed other models in both the testing and training phases. A significant downward in the volume of water flowing through the region's rivers and streams in the next decades. It is critical to enhance climate‐resilient water infrastructure financing, establish an early warning system for drought, introduce best management practices, implement integrated water resource management, public awareness, and support water research to alleviate the negative consequences of drought and increase resilience against the effects of climate change on Turkey's water resources.