Analysing Mental Health Trends and Predictors Among Students using Reinforcement Learning

Aim: The aim of the study is to analyse and improve the efficiency of mental health prediction among students by using Reinforcement Learning (RL) and comparing its performance with the K-Nearest Neighbours (KNN) algorithm. The focus of this project is to enhance prediction accuracy and efficiency using collection student data. Materials and Methods: In this research there are two groups. Group 1 refers to the K-Nearest Neighbours model with 200 samples and Group 2 refers to the Reinforcement Learning algorithm with 250 samples. The G Power value is 80% and threshold is 0.05% with a confidence interval of 95%. Result: The Reinforcement Learning model has significantly higher efficiency and prediction accuracy than KNN algorithm. Reinforcement Learning achieved prediction accuracy ranging from 85% to 94% and the K-Nearest Neighbour achieved prediction accuracy ranging from 72% to 81%. The output frequency for maximum efficiency in the Reinforcement Learning algorithm with a significance of about 0.00065. Conclusion: In this work, it is observed that the Reinforcement Learning algorithm has significantly better efficiency than the K-Nearest Neighbour Algorithm.