NEURAL NETWORK MODEL FOR PREDICTING STUDENTS’ ACADEMIC PERFORMANCE
Authors:
1.
Kazan National Research Technical University named after A.N. Tupolev
2.
Kazan National Research Technical University named after A.N. Tupolev
Type:
Article
Pages:
from 137
to 145
Status:
Published
Received:
20.01.2026
Accepted:
02.02.2026
Published:
26.02.2026
Language:
Russian
Keywords:
MACHINE LEARNING, NEURAL NETWORKS, ACADEMIC PERFORMANCE PREDICTION, EDUCATIONAL DATA, NEURAL NETWORK MODEL, ACADEMIC ACHIEVEMENT
Abstract:
This article examines the application of machine learning and neural network methods to predicting students' academic performance based on the analysis of educational data. The relevance of the study is driven by the growing volume of digital data in modern educational systems and the need to develop intelligent decision support tools aimed at improving the quality of education and early identification of students at academic risk. The object of the study is data characterizing students' academic and extracurricular activities, including attendance indicators, study load, socio-demographic characteristics, and previous learning outcomes. The subject of the study is the process of predicting students' academic performance using neural network classification models. The paper analyzes existing forecasting methods used in educational analytics and substantiates the choice of a multilayer fully connected neural network as the most suitable tool for processing tabular data. The model is implemented using the scikit-learn library. Particular attention is paid to the stages of data preprocessing, including gap handling, coding of categorical features, scaling of numerical parameters, and the formation of target variable classes. In the experimental studies, the performance of the neural network model was compared with different numbers of output classes of the target variable. Performance was assessed using classification metrics such as Accuracy, Precision, Recall, and F1-score. The results showed that the three-class classification provides an optimal balance between prediction accuracy and the practical interpretability of the results. The neural network model may be useful for predicting academic performance at various stages of learning.
This article examines the application of machine learning and neural network methods to predicting students' academic performance based on the analysis of educational data. The relevance of the study is driven by the growing volume of digital data in modern educational systems and the need to develop intelligent decision support tools aimed at improving the quality of education and early identification of students at academic risk. The object of the study is data characterizing students' academic and extracurricular activities, including attendance indicators, study load, socio-demographic characteristics, and previous learning outcomes. The subject of the study is the process of predicting students' academic performance using neural network classification models. The paper analyzes existing forecasting methods used in educational analytics and substantiates the choice of a multilayer fully connected neural network as the most suitable tool for processing tabular data. The model is implemented using the scikit-learn library. Particular attention is paid to the stages of data preprocessing, including gap handling, coding of categorical features, scaling of numerical parameters, and the formation of target variable classes. In the experimental studies, the performance of the neural network model was compared with different numbers of output classes of the target variable. Performance was assessed using classification metrics such as Accuracy, Precision, Recall, and F1-score. The results showed that the three-class classification provides an optimal balance between prediction accuracy and the practical interpretability of the results. The neural network model may be useful for predicting academic performance at various stages of learning.
Keywords:
MACHINE LEARNING, NEURAL NETWORKS, ACADEMIC PERFORMANCE PREDICTION, EDUCATIONAL DATA, NEURAL NETWORK MODEL, ACADEMIC ACHIEVEMENT
MACHINE LEARNING, NEURAL NETWORKS, ACADEMIC PERFORMANCE PREDICTION, EDUCATIONAL DATA, NEURAL NETWORK MODEL, ACADEMIC ACHIEVEMENT
