Suicidal Tendency Neural Identifier in University Students from Aguascalientes, Mexico

This document describes the increasingly frequent phenomenon of suicide among university students; as well as some of its effects in the educational environment and the need to develop tools that facilitate the application of public policies for its attention and follow-up. As a result of this research, features that have the greatest impact on suicidal behavior were identified and deep neural network was designed for the classification of young people with these tendencies. Some of the results of this research suggest that bipolar disorder, geographic location and drugs addiction are some of the critical factors to be consider in such a problem. The neural network used, is a dense deep model built with 3 hidden layers and its precision is greater than 90%. With the use of tools such as the one presented; this group of researchers believes it is feasible to intervene to reduce this phenomenon.