نوع مقاله : مقاله پژوهشی
نویسنده
دانشکده مهندسی، مرکز آموزش عالی محلات
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسنده [English]
Predicting students' academic success has always been one of the significant challenges in educational systems. Numerous studies indicate a meaningful relationship between students' lifestyle and their academic performance. The use of artificial intelligence methods can serve as a powerful tool for analyzing data related to academic behaviors and lifestyle and predicting academic success. This research aimed to develop a system for predicting students' academic success based on lifestyle indicators using machine learning algorithms. After data collection, preprocessing steps including normalization and feature selection were performed using the ReliefF algorithm. Subsequently, predictive models were developed using four methods: Random Forest (RF), Multi Layer Perceptron (MLP) , K-Nearest Neighbors (KNN) and Linear Regression(LR). The results showed that the LR model, after a 50% reduction in features and the selection of the most effective indicators such as daily study hours, mental health status, and sleep hours, with values of MAE=4.33 and RMSE=5.42, demonstrated the best efficiency from time and performance measures aspects. This model can be utilized in designing intelligent educational platforms. The findings confirm that analyzing lifestyle data using machine learning methods can create prediction systems with acceptable accuracy, although expanding the statistical sample and adding some other features could improve results in future research.
کلیدواژهها [English]
)