Exploring Student Achievement through Deep Learning: Uncovering the Dynamic Interactions Between Inclusion, Environment, and Academic Success

Abstract

This study focuses on the use of deep learning techniques to predict student performance based on classroom participation and learning environment. With the continuous growth of data-driven learning, understanding the original factors that affect academic achievement has become important. Predictive models often fail to capture the interactions between classroom participation, environment, and their effects on student achievement. This research aims to provide more accurate and predictive models by integrating deep learning techniques. The collects data on 1,000 students, including classroom participation (attendance, engagement) and environment (class activities). The dataset is prepared for analysis through preprocessing steps such as standardization and underestimation. A deep learning model was developed using Tensor Flow, which consists of an input layer, a hidden layer with Relook boosting, and an output layer with linear functions. The model was trained using the Adam optimizer, and mean squared error (MSE) and R-squared (R²) metrics were used to analyze its performance. The results showed that there was a positive correlation between class participation and student performance, with participation contributing 65% to the model's accuracy. Environmental factors, although small, accounted for 35% of the estimated energy. The deep learning model has an MSE of 0.35 and an R² score of 0.92, indicating that it captures the variance in student performance well. The results revealed a positive correlation between prediction and actual performance, with students who interacted socially performing better regardless of the classroom. For example, students with fewer social connections are more likely to be influenced by their environment. This study not only supports research on prediction in education, but also provides valuable insights for teachers. By identifying key determinants of success, such as engagement and classroom environment, interventions can be designed to support at-risk students. The study also shows the potential of deep learning models in creating personalized learning experiences and improving resource allocation in education. Therefore, this study will improve learning outcomes and create higher quality education.

Keywords: Student achievement, Deep learning, Educational engagement, Learning environment, Academic achievement, Dynamic interaction