A Machine Learning Approach to Detecting Suicidal Tendencies in Adolescents

Abstract

Nowadays suicide has become a serious crime. People commit suicide for many reasons. Due to the advancement of social media, people post various types of posts before committing suicide. Understanding the environmental risk factors that affect suicide thoughts and behavior throughout time will be greatly aided by this study. We will collect various types of data from various online platforms and identify them with the help of machine learning models. We use some algorithms to find out the best accuracy. To find out the best accurate results we implement classifier such as Naive Bayes, Support Vector Machine (SVM), Logistic Regression, Decision Tree Classification, Random Forest and Gradient Boosting Classifier. The models were evaluated based on their accuracy, precision, recall and F-1 score. Naive Bayes had the lowest accuracy at 73%, while Support Vector Machine (SVM) secured the top positions with 92.32% accuracy. This research highlights the importance of integrating advanced machine learning techniques into mental health care to facilitate early intervention and support for at-risk adolescents. By leveraging technology, we can enhance the effectiveness of suicide prevention strategies and ultimately save lives.