Predicting the insulating paper state of the power transformer based on XGBoost/LightGBM models DOI Creative Commons
Sherif S. M. Ghoneim, Mohammed Baz,

Ali Alzaed

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Май 22, 2025

Power transformer plays a crucial role in the power networks. Most of malfunctions was due to failure insulating systems. The utilities keen on contentious operation network, so early detection faults avoiding undesirable outage from service. paper state is an indication health and aging it may lead transformer, some periodic routine test must be performed insulting oil get information about condition. value degree polymerization (DP) key state. Various recommended tests such as dissolved gases (DGA), breakdown voltage (BDV), interfacial tension (IF), acidity (ACI), moisture content (MC), color (OC), dielectric loss (Tan δ), furans concentration specifically (2-furfuraldehyde (2-FAL)) were carried out correlate between these variables DP then collected data used supply XGBoost/LightGBM build artificial intelligence model predict results indicated that great ability proposed with high accuracy. Of various configurations for two classification models, one achieved perfect prediction accuracy (100%) other showed values 1.0 down 0.955.

Язык: Английский

Accuracy Comparison of Machine Learning Algorithms on World Happiness Index Data DOI Creative Commons
Sadullah Çelik, Bilge DOĞANLI, Mahmut Ünsal Şaşmaz

и другие.

Mathematics, Год журнала: 2025, Номер 13(7), С. 1176 - 1176

Опубликована: Апрель 2, 2025

This study aims to compare the accuracy performances of different machine learning algorithms (Logistic Regression, Decision Tree, Support Vector Machines (SVMs), Random Forest, Artificial Neural Network, and XGBoost) using World Happiness Index data. The is based on 2024 Report data employs indicators such as Ladder Score, GDP Per Capita, Social Support, Healthy Life Expectancy, Freedom Determine Choices, Generosity, Perception Corruption. Initially, K-Means clustering algorithm applied group countries into four main clusters representing distinct happiness levels their socioeconomic profiles. Subsequently, classification are used predict cluster membership scores obtained serve an indirect measure quality. As a result analysis, Logistic SVM, Network achieve high rates 86.2%, whereas XGBoost exhibits lowest performance at 79.3%. Furthermore, practical implications these findings significant, they provide policymakers with actionable insights develop targeted strategies for enhancing national improving well-being. In conclusion, this offers valuable information more effective analysis by comparing various algorithms.

Язык: Английский

Процитировано

0

A new approach for predicting academic performance DOI Open Access
Abdallah Maiti, Abdallah Abarda, Mohamed Hanini

и другие.

Procedia Computer Science, Год журнала: 2025, Номер 257, С. 1257 - 1262

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

Predicting the insulating paper state of the power transformer based on XGBoost/LightGBM models DOI Creative Commons
Sherif S. M. Ghoneim, Mohammed Baz,

Ali Alzaed

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Май 22, 2025

Power transformer plays a crucial role in the power networks. Most of malfunctions was due to failure insulating systems. The utilities keen on contentious operation network, so early detection faults avoiding undesirable outage from service. paper state is an indication health and aging it may lead transformer, some periodic routine test must be performed insulting oil get information about condition. value degree polymerization (DP) key state. Various recommended tests such as dissolved gases (DGA), breakdown voltage (BDV), interfacial tension (IF), acidity (ACI), moisture content (MC), color (OC), dielectric loss (Tan δ), furans concentration specifically (2-furfuraldehyde (2-FAL)) were carried out correlate between these variables DP then collected data used supply XGBoost/LightGBM build artificial intelligence model predict results indicated that great ability proposed with high accuracy. Of various configurations for two classification models, one achieved perfect prediction accuracy (100%) other showed values 1.0 down 0.955.

Язык: Английский

Процитировано

0