Deep Learning and Natural Language Processing-Based Model for the Prediction of Suicidal Ideation in Military Personnel

The problem of people having thoughts of ending their own lives is becoming more widespread throughout society. The rise in people’s levels of stress, anxiety, sadness, and sleeplessness, as well as a host of other mental health issues, is the root cause of suicidal thinking. The most prevalent contributing causes to an increase in suicidal thoughts are feelings of depression and stress. Therefore, depending to the circumstances, members of the armed forces experience higher levels of psychological stress and have a larger probability of attempting suicide than the general population does. Suicidal thoughts may be brought on by extreme mental stress, which is a key factor in the rise of suicide attempts. In order to determine whether or not a person is having suicidal thoughts, this article makes use of the deep learning approach known as multilayer perceptron. After applying the aforementioned algorithms, a ROC curve and a precision-recall curve were used in order to determine the algorithm’s level of accuracy. The Cohen Kappa was used to the confusion matrix in order to determine the model’s level of accuracy.