Machine Learning Approach to Detect the Anxiety and Depression Level of Video Game Dependent Individuals

The most prominent kind of entertainment in the present day is playing video games, which are experienced by players of all ages. When a player is fully engaged in a game, they are utterly unaware of their surroundings. The primary objective of this study is to prove the close relationship between excessive gaming and the negative effects of anxiety and depression. It also looks at how well machine learning techniques can be used to predict the level of anxiety and depression in people who have become dependent on video games. In order to predict anxiety and depression, previous research has mostly used cross-sectional data and standard-scale questionnaires. We collected information from an online survey and utilized the use of several machine-learning approaches in order to predict the anxiety and depression that persons who are dependent on video games will certainly suffer. Furthermore, we employed feature selection techniques to identify features that were most effective at predicting both anxiety and depression. When taking only into account the most important factors, the SVM classifier obtained a depression prediction accuracy of 93.48% whereas the k-NN classifier achieved an accuracy of 91.44% for anxiety. The study's findings offer a promising start by using machine learning to anticipate anxiety and depression in those who depend on video games.