AI-Driven Insights into the Relationship Between Psychological State and Academic Performance

This paper presents an experiment on the application of AI, specifically machine learning techniques, to advance psychological research and practice. The study investigates the relationship between the psychological state of computer scienc students and their academic performance. Using a guided survey, a local dataset was created with 265 students and 31 variables. This dataset was then used to train AI models aimed at predicting academic performance based on selected input factors. While some traditional machine learning algorithms demonstrated relatively low accuracy, the application of the Random Forest algorithm with SMOTE(Synthetic Minority Over-sampling Technique) significantly improved predictive accuracy. These promising results highlight the potential of AI to support and accelerate research in psychology, enhancing both study precision and experimental efficacy.