Date of Award

7-2020

Document Type

Project

Degree Name

Master of Science in Information Systems and Technology

Department

Information and Decision Sciences

First Reader/Committee Chair

Benjamin Becerra

Abstract

Suicide is one of the leading health concerns in United States among adolescents and the presence of suicidal ideation (SI) is quite high, with ~20-30% of adolescents reporting it at some point. Though we have seen growth and development in the prevention of suicide, there is limited research on the ability to identify the adolescents which might be at risk for SI. The objective behind the project is to identify adolescents with SI using machine learning.

The project shows statistics from different articles on adolescents in the U.S. For this study, adolescent data was taken from NSDUH 2018. Moreover, detailed associations between demographics, mental health features, etc. and SI were conducted.

From the analysis we saw that functional impairment during MDE on social life, family relationships, schoolwork and home chores play an important role in identifying SI among adolescents. Moreover, machine learning algorithm was implemented to predict SI (accuracy score ~65-70%).

This paper also shows the importance of each feature on the prediction of SI among adolescents. In future work these features could be used to develop an application which will predict SI among adolescents. This screening tool could be potentially utilized by healthcare professionals to screen adolescents during patient encounters. In the future more data could be combined from different years and countries, including the social media information as well.

Share

COinS