Classification Tree Technique Research Articles

Machine learning based classifiers for dynamic and transient disturbance classification in smart microgrid system

The smart microgrid system should have the ability to rapidly detect and classify every type of disturbance that happens in the network to operate the protection scheme and maintain the power quality. Both dynamic and transient types of disturbances are considered in this study, and the classification of each type of disturbance has been done using different machine learning-based techniques. A novel ensemble classifier has been proposed after comparing the performance of the decision tree (DT), k nearest neighbor (KNN), support vector machine (SVM), and ensemble classifier. The performances of all the classifiers have been evaluated in terms of accuracy, sensitivity, confusion matrices, and receiver operating characteristics (ROC). A very satisfactory result has been obtained in the ensemble tree classification technique with 99.3% accuracy in 2.1 s of training time. Therefore, it can be suggested to identify and classify every transient as well as steady-state events of the microgrid network.

Potential suicide risk among the college student population: machine learning approaches for identifying predictors and different students’ risk profiles

BackgroundSuicide is one of the leading causes of death among young people and university students. Research has identified numerous socio-demographic, relational, and clinical factors as potential predictors of suicide risk, and machine learning techniques have emerged as promising ways to improve risk assessment.ObjectiveThis cross-sectional observational study aimed at identifying predictors and college student profiles associated with suicide risk through a machine learning approach.MethodsA total of 3102 students were surveyed regarding potential suicide risk, socio-demographic characteristics, academic career, and physical/mental health and well-being. The classification tree technique and the multiple correspondence analysis were applied to define students’ profiles in terms of suicide risk and to detect the main predictors of such a risk.ResultsAmong the participating students, 7% showed high potential suicide risk and 3.8% had a history of suicide attempts. Psychological distress and use of alcohol/substance were prominent predictors of suicide risk contributing to define the profile of high risk of suicide: students with significant psychological distress, and with medium/high-risk use of alcohol and psychoactive substances. Conversely, low psychological distress and low-risk use of alcohol and substances, together with religious practice, represented the profile of students with low risk of suicide.ConclusionsMachine learning techniques could hold promise for assessing suicide risk in college students, potentially leading to the development of more effective prevention programs. These programs should address both risk and protective factors and be tailored to students’ needs and to the different categories of risk.

Assessment of satisfaction in emergency remote learning from pharmacy students’ perspective in a public university in Brazil

The remote teaching-learning process in the Baccalaureate of Pharmacy at UFMG brought many challenges to the students, thus our objective was to assess the satisfaction of the students with the ERL. It was evaluated when the classes were instituted by means of Emergency Remote Learning (ERL), during the coronavirus pandemic (SARS-CoV-2). A questionnaire addressing teaching-learning data, sociodemographic habits and lifestyle was collected through a virtual platform and, data are analyzed by linear regression and by the Classification And Regression Tree (CART) technique. The overall mean students´ satisfaction with ERL was 3.12±1.10 (Likert scale 1-5), presenting an inverse correlation with learning in the professional cycle of the course (p<0.001). Only 3.1% indicated 'none or little learned', variable classified as maximum satisfaction. Low satisfaction was registered with the quality of the practical classes (3.06±1.22) revealing the limitation of ERL in the case of a lot of practical content, as for Pharmacy course. A high level of satisfaction was related to the didactics and methodology of the professors (3.55±0.99), showing their commitment and dedication to the ERL, despite the difficulties of its implementation. The study results point for improvements in content, course organization, or the use of different technologies to enhance student satisfaction with remote learning format.

Utilizing Extremely Fast Decision Tree (EFDT) Algorithm to Categorize Conflict Flow on a Software-Defined Network (SDN) Controller

Software-Defined Networks (SDNs) provide a contemporary approach to networking technology, offering a versatile and dynamically efficient network architecture for enhanced surveillance and performance. However, SDN architectures may encounter flow conflicts. These conflicts arise when modifications are made to specific flow properties, such as priority, match field, and action. Despite the existence of recommended solutions, the process of resolving conflicts in SDN continues to encounter difficulties. This study proposes an Extremely Fast Decision Tree (EFDT) classification technique to detect and categorize conflicts inside the flow table. The novelty of this method is based on the development of an accurate and effective machine-learning technique implemented on the Ryu controller plane and validated using the Mininet simulator. The effectiveness and efficiency of the proposed method were evaluated using various indicators, demonstrating superior performance in recognizing and categorizing conflict flow types in all flow sizes ranging from 10,000 to 100,000.

Unveiling IoT Customer Behaviour: Segmentation and Insights for Enhanced IoT-CRM Strategies: A Real Case Study.

In today's competitive landscape, achieving customer-centricity is paramount for the sustainable growth and success of organisations. This research is dedicated to understanding customer preferences in the context of the Internet of things (IoT) and employs a two-part modeling approach tailored to this digital era. In the first phase, we leverage the power of the self-organizing map (SOM) algorithm to segment IoT customers based on their connected device usage patterns. This segmentation approach reveals three distinct customer clusters, with the second cluster demonstrating the highest propensity for IoT device adoption and usage. In the second phase, we introduce a robust decision tree methodology designed to prioritize various factors influencing customer satisfaction in the IoT ecosystem. We employ the classification and regression tree (CART) technique to analyze 17 key questions that assess the significance of factors impacting IoT device purchase decisions. By aligning these factors with the identified IoT customer clusters, we gain profound insights into customer behaviour and preferences in the rapidly evolving world of connected devices. This comprehensive analysis delves into the factors contributing to customer retention in the IoT space, with a strong emphasis on crafting logical marketing strategies, enhancing customer satisfaction, and fostering customer loyalty in the digital realm. Our research methodology involves surveys and questionnaires distributed to 207 IoT users, categorizing them into three distinct IoT customer groups. Leveraging analytical statistical methods, regression analysis, and IoT-specific tools and software, this study rigorously evaluates the factors influencing IoT device purchases. Importantly, this approach not only effectively clusters the IoT customer relationship management (IoT-CRM) dataset but also provides valuable visualisations that are essential for understanding the complex dynamics of the IoT customer landscape. Our findings underscore the critical role of logical marketing strategies, customer satisfaction, and customer loyalty in enhancing customer retention in the IoT era. This research offers a significant contribution to businesses seeking to optimize their IoT-CRM strategies and capitalize on the opportunities presented by the IoT ecosystem.

Can predictive factors determine the time to treatment initiation for oral and oropharyngeal cancer? A classification and regression tree analysis.

This ecological study aimed to identify the factors with the greatest power to discriminate the proportion of oral and oropharyngeal cancer (OOC) records with time to treatment initiation (TTI) within 30 days of diagnosis in Brazilian municipalities. A descriptive analysis was performed on the variables grouped into five dimensions related to patient characteristics, access to health services, support for cancer diagnosis, human resources, and socioeconomic characteristics of 3,218 Brazilian municipalities that registered at least one case of OOC in 2019. The Classification and Regression Trees (CART) technique was adopted to identify the explanatory variables with greater discriminatory power for the TTI response variable. There was a higher median percentage of records in the age group of 60 years or older. The median percentage of records with stage III and IV of the disease was 46.97%, and of records with chemotherapy, radiation, or both as the first treatment was 50%. The median percentage of people with private dental and health insurance was low. Up to 75% had no cancer diagnostic support services, and up to 50% of the municipalities had no specialist dentists. Most municipalities (49.4%) started treatment after more than 30 days. In the CART analysis, treatment with chemotherapy, radiotherapy, or both explained the highest TTI in all municipalities, and it was the most relevant for predicting TTI. The final model also included anatomical sites in the oral cavity and oropharynx and the number of computed tomography services per 100,000. There is a need to expand the availability of oncology services and human resources specialized in diagnosing and treating OOC in Brazilian municipalities for a timely TTI of OOC.

Pos Aja Customer Segmentation at the Yogyakarta Post Office Using the RFM (Recency, Frequency, Monetary) Method

Every business must continue to innovate due to the industry's rapid evolution if it hopes to avoid being disrupted by innovations in the future. The Pos Aja! app, a revolutionary new offering from courier and logistics services provider Pos Indonesia, was introduced in 2020. With Just Post!, consumers can transact from anywhere without physically visiting the post office. Pos Aja! facilitates not just transaction processing for clients but also helps businesses better understand their clientele. The technique known as customer relationship management, or CRM, works to maintain and improve ties between businesses and their clients. Client segmentation is one method of investigating client profiles. Recency, Frequency, and Monetary analysis, or RFM, is the customer segmentation method employed in this study in order to identify the traits of Pos Aja! customers. Three types of Pos Aja! customers—Gold, with 857 clients; Silver, with 566 customers; and Bronze, with 688 customers—were obtained based on the RFM data. The RFM data were then sorted to determine the category of new Pos Aja! clients. The Random Forest, Decision Tree, and Support Vector Machine (SVM) classification techniques are employed. With an accuracy level of 96.69%, the SVM model emerged as the top performer among these three models. Therefore, the SVM model can be applied to determine new client groups.

PREDICTION OF BODY WEIGHT OF NIGERIAN NON-DESCRIPT GOATS FROM MORPHOMETRIC TRAITS USING CLASSIFICATION AND REGRESSION TREE MODEL

This study was conducted to evaluate the relationship between body weight and eleven (11) morphometric traits (body weight, body length, height at withers, rump height, chest girth, hind leg, fore leg, head length, ear length, neck length, and tail length) of non-descript goats using classification and regression tree technique. The data were generated from 120 non-descript goats randomly selected from different herds in three LGA areas of Kaduna State, North West Nigeria. Pearson’s moment correlation (r) between body weight and morphometric traits ranged from low to high values (r = 0.21-0.97; P≤0.05, P≤0.01). Based on the importance of the independent variables in predicting the body weight of goats, six body measurements namely; chest girth, body length, rump height, height at withers, head length and neck length were found to be more efficient. Thus, they were the variables entered to obtain the optimal regression tree. Among these six variables, chest circumference was found to be the primary splitting variable; and together with neck length accounted for about 84.20% of the variation in body weight. The regression tree analysis indicated that animals with chest circumference > 60.00cm and neck length >15cm could be expected to have higher body weights. This information could be exploited by livestock producers and researchers for determining the feed amount, drug dose, and market price of an animal, management, selection and genetic improvement of Nigerian non-descript goats.

A novel diagnosis method for schizophrenia based on globus pallidus data

This research aims to diagnose schizophrenia with machine learning-based algorithms. Bayesian neural network, logistic regression, decision tree, k-nearest neighbor, and gaussian kernel classification techniques are investigated to diagnose schizophrenia with data from 125 persons. This study showed that left lateral ventricles and left globus pallidus volumes and their percentages in the brain were significantly lower than HCs in FEP patients. Using brain volumes, we were able to diagnose FEP with an accuracy of 73.6 % via logistic regression and with an accuracy of 86.4 % using the SVM kernel classifier method. Therefore, brain volumes can be used to diagnose FEP with the SVM kernel classifier method.

Application of classification trees technique in optimization of parameters of production of film-coated tablets

Application of classification trees technique in optimization of parameters of production of film-coated tablets

Impact analysis and optimization of notched high-gain Antipodal Vivaldi antenna using machine learning

Optimizing antenna parameters without knowing their impact on the gain is a time-consuming and iterative process. This article shows how the Classification and Regression Tree (CART) technique, the Random Forest (RF) algorithm, and the Principal Component Analysis (PCA) method is used to study the effect of geometrical parameters of rectangular notches of the Antipodal Vivaldi antenna on gain. Carving out the notches is a general approach for end-fire antenna design. Integrating the EM simulation with machine learning could reduce the time spent on AVA design. In this approach, the gain category of the antenna under design: Low, Medium, and High, is first classified by the three multiclass machine learning models, K-Nearest Neighbor (KNN), Artificial Neural Network (ANN), and Support Vector Machine (SVM). When the predicted category of gain belongs to the High class, in that case, the four regression models, Polynomial, SVM, ANN, and RF, predict the gain value to anticipate the antenna gain according to the notch dimensions to reduce the optimization time of the simulation-driven antenna. In this work, the band-limited approach of notches on gain is also been validated using simulation before fabrication. Machine learning models are optimized the antenna parameters to achieve a high gain within the frequency range of 20 to 40 GHz. A peak simulated gain of 19 dB is achieved at the frequency of 27 GHz. The proposed structure is fabricated utilizing a single-layer Rogers substrate (RO4003C) with a dielectric constant of 3.38 and compact dimensions of 96 mm × 174 mm × 0.508 mm, i.e., 0.32 × 0.58 × 0.0017 λ03; here the symbol λ0 denotes the wavelength of free space at the minimum operating frequency.

Profiles of University Students Who Graduate on Time: A Cohort Study from the Chilean Context.

Graduating from higher education on time is an important topic, given its relation to students' academic success and the efficiency of the institutions. However, a low percentage of university students finish their studies on time, which poses a challenge that requires the identification of the factors that account for this phenomenon. This study aimed to identify and characterize profiles of students who graduate on time. The population is 514 university students (45% men, 55% women), with an average age of 19.5 years (SD= 1.9) studying business at a university in Chile who belong to four cohorts entering between 2011 and 2014. The results obtained from the Classification and Regression Tree (CART) technique demonstrate eight student profiles constructed considering different variables at the pre-university, transition-motivation, and university levels. As the primary outcome, the profile of the student who graduates on time is characterized by a good performance in the first year of university, enters university right after high school, and takes advantage of institutional support by participating in academic tutoring. These findings suggest that institutions can implement specific strategies from the beginning of the university journey to promote on-time graduation.

LICENSING AND ENFORCEMENTIN MUNICIPAL ENVIRONMENTAL DEPARTMENTS IN THE STATE OF RIO GRANDE DO SUL, BRAZIL

This study evaluated the environmental departments of some municipalities in Rio Grande do Sul, Brazil, regarding their structure and licensing and enforcement processes. Information was collected through online questionnaires sent to municipalities. The questionnaires addressed the staff, the training of the technical staff and the work structure of the environmental departments in the municipalities. Evaluations and classifications were conducted using the classification tree technique based on municipal data on the environmental licenses issued, and additionally the infraction notices issued as response variables. Thus, it was observed that human resources, effective civil servants and the work team composition are the variables that best explain the issuance behavior of the municipal environmental licenses, and in parts the infraction notices. The other variable that most satisfactorily explains the issuance of infraction notices in the municipalities is the use of their own website with environmental software in the environment departments.

Classification trees as machine learning tool to explore consumers’ purchasing decision pathway. A case-study on parent’s perception of baby food jars

Classification and regression trees are machine learning methods useful to model non-linear and complex relationships. The objective of this study was to use classification tree method to model purchasing parents’ decision of buying or rejecting baby food jars and to identify the different pathways in the decision of consumers that explain the different purchasing patterns. For that, parents (n = 100) of babies (6 to 36 months old) evaluated 16 baby food jar images. For each one, parents indicated their purchase intention but also evaluated other aspects like expected liking, trust, health, and nutritive perception. Three clusters with different purchasing patterns were obtained and classification trees showed the different hierarchy of factors in the pathway toward the purchasing decision. For cluster 1 decision was based mainly on liking but conditioned to trust perception. For cluster 2, the decision was firstly based on healthiness perception and conditioned to liking. Cluster 3, the decision included a more complex pathway initiated by trust and followed by liking, nutritive and healthiness perceptions. The approach combining clustering and classification tree techniques is useful to explore and understand individual differences in consumers purchasing decision.

Identity profile of young people experiencing a sense of risk on the internet: A data mining application of decision tree with CHAID algorithm

Social media, interactive games, and digital platforms are just some of the ways in which the internet has become part of our lives, and more specifically so among young people. Among other things, this technology is the reason young people are spending increasingly more time online than offline, prompting numerous risks that they are sometimes unaware of. The following quantitative research is based on this premise, and it pursues two objectives, namely, to discover whether or not young people experience a sense of risk linked to their use of the internet and identify the kind of young people that say they do or do not perceive any such risk. A total of 1991 young people answered the Technology Use Studies questionnaire. The classification tree technique based on the Chi-square Automatic Interaction Detector (CHAID) was applied to identify the predictive variables associated with the sense of risk in young people. Two highly significant results have been forthcoming: firstly, we present the factors with the biggest impact on a young person's internet-related sense of risk, involving isolation when they cannot go online, followed by variables such as gender, the use of TikTok, and videogames, the consumption of pornography, and distraction from study time. Secondly, our results indicate that the young people that feel isolated when they cannot go online are precisely the ones that do not experience a sense of online risk. It is therefore of vital importance for education to generate and provide young people with strategies that allow them to establish a healthy and responsible relationship with technologies such as the internet.

A decision tree-based damage estimation approach for preliminary seismic assessment of reinforced concrete buildings

This study aims to introduce an earthquake-induced damage classification approach for seismic vulnerability assessment of reinforced concrete buildings. Through the use of the damage data collected from post-earthquake inspections after the 2003 Bingöl Earthquake in Turkey, two models were constructed by the decision tree classification technique considering nine building-specific features as the estimation variables in the analysis. The first model was developed for the prediction of the observed damage states of the buildings, whereas the second one concerning the life safety level assessment, was proposed for distinguishing the extremely vulnerable buildings for seismic prioritization. In the validation process, the leave-one-out cross validation technique was adopted to deal with the small sample size of the building inventory. Among the estimation variables, the priority index and the existence of short columns were found to have the highest importance in classification. Results have revealed that the proposed model for life safety level assessment was capable of discriminating the cluster of severely damaged and collapsed buildings from the entire database with an accuracy of 70.59%. Hence, the damage classification approach adopted in this study has the potential for improving effective tools for seismic risk assessment of the existing buildings.

공급사슬 혁신지향과 기업성과의 관계 : 분류나무 분석

Purpose This study empirically analyzes the impact of innovation orientation on business performances in supply chain Methods Diverse supply chain workers answered the survey using a Likert scale measure. The innovation orientation of supply chain is classified into the dimensions of organization, exploration, and exploitation, whereas business performances are measured by logistics efficiency and financial performance. By applying classification tree technique, several interactive effects of innovation orientation are identified Results The analysis results indicate that many variables of innovation orientation are crucial for improving logistics efficiency as well as financial success for corporations of supply chain. By pairing variables selected for the first split of the trees and the ones for the second split, we attempt to suggest how organizational, explorative, and exploitative innovation can contribute to overall performance of supply chain Conclusion The findings of this study provide valuable empirical lessons about the importance of innovation to all participants of supply chain together with supportive academic literature

Discovery of Intentional Self-Harm Patterns from Suicide and Self-Harm Surveillance Reports.

The purpose of this study was to identify patterns of self-harm risk factors from suicide and self-harm surveillance reports in Thailand. This study analyzed data from suicide and self-harm surveillance reports submitted to Khon Kaen Rajanagarindra Psychiatric Hospital, Thailand. The process of identifying patterns of self-harm risk factors involved: data preprocessing (namely, data preparation and cleaning, missing data management using listwise deletion and expectation-maximization techniques, subgrouping factors, determining the target factors, and data correlation for learning); classifying the risk of self-harm (severe or mild) using 10-fold cross-validation with the support vector machine, random forest, multilayer perceptron, decision tree, k-nearest neighbors, and ensemble techniques; data filtering; identifying patterns of self-harm risk factors using 10-fold cross-validation with the classification and regression trees (CART) technique; and evaluating patterns of self-harm risk factors. The random forest technique was most accurate for classifying the risk of self-harm, with specificity, sensitivity, and F-score of 92.84%, 93.12%, and 91.46%, respectively. The CART technique was able to identify 53 patterns of self-harm risk, consisting of 16 severe self-harm risk patterns and 37 mild self-harm risk patterns, with an accuracy of 92.85%. In addition, we discovered that the type of hospital was a new risk factor for severe selfharm. The procedure presented herein could identify patterns of risk factors from self-harm and assist psychiatrists in making decisions related to self-harm among patients visiting hospitals in Thailand.

Disease Detection System in Tobacco Leaves based on Edge Detection with Decision Tree Classification Method

Indonesia has long been known as an agricultural country, one of Indonesia's best agricultural products is tobacco. Tobacco with good quality can be seen from the leaves, in fact the diseases that attack tobacco are of various types which can be seen from the changes in the tobacco leaves starting from the seeding and planting period. In the tobacco growing period, it is divided into two major parts, namely seeding period and the planting period, so that diseases that attack tobacco are also divided into two, namely diseases that attack during seeding and planting. This research is limited to diseases that attack tobacco at the time of planting, because at the time of seedling the tobacco has not yet produced leaves. When tobacco enters the planting period, at this time tobacco leaves begin to form. Good care is needed at this time such as fertilization, nutrition, vitamins, and pest control in order to obtain healthy tobacco so that tobacco is not susceptible to disease. Tobacco that lacks nutritional intake will be susceptible to diseases including fungi, bacteria, and viruses. The disease attack on tobacco has its own characteristics that appear on tobacco leaves. Early detection of the disease is very important so that disease control can be precise and the spread of the disease can be prevented so as not to cause endemic. In this research, an early detection system of tobacco leaf disease based on image processing will be designed. Normalization image, grayscale technique, folllowed by edge detection will be applied in these image so that from here the entropy, energy, and inertia values of the image can be obtained using statistical measures, and the last one using the decision tree classification technique can be classified as uninfected leaf or infected leaf. In this study, feature extraction from images of tobacco leaves that are not infected with the virus using grayscale techniques followed by edge detection produces an average statistical measure with entropy (h) values between 2,341 to 2,676, energy (e) values between 6,112 to 6,665, and inertia values. (i) between 3,322 to 3,576, while for leaves infected with the virus the average value of entropy (h) is between 4,543 to 5,576, the average value of energy (e) is between 12,212 to 13,455, and the average value of inertia (i) between 5,343 to 6,597.

(Retracted) Hurricane damage assessment in satellite images using hybrid VGG16 model

Hurricanes are one of the most disastrous natural phenomena occurring on Earth that cause loss of human lives and immense damage to property. A damage assessment method has been proposed for damage caused to buildings due to Hurricane Harvey that hit the Texas region in the year 2017. The aim of our study is to predict if there is any damage to the buildings present in the postdisaster satellite images. Principal component analysis has been used for the visualization of data. The VGG16 model has been used for extracting features from the input images. K-nearest neighbor (KNN), logistic regression, decision tree, random forest, and XGBoost classification techniques have been used for classification of the images whose features have been extracted from VGG16. Best accuracy of 97% is obtained by KNN classifier for the balanced test set, and accuracy of 96% is obtained by logistic regression for the unbalanced test set.