Predictive Model of Suicidal Ideation among American Adult for Suicide Risk Management and Economic Cost Reduction

Suicide is a national public health problem in modern society. Early detection and prevention of suicide thoughts needs be addressed to save people’s lives and reduce economic productivity loss. This research applies the machine learning modelling approach to suicidal ideation prediction using unexplored data from the 2019 National Survey on Drug Use and Health (NSDUH). Significant risk factors of depression disorder, chronic mental illness, substances abuse, and unmet medical service needs are identified for the suicidal ideation predictive modelling. The models use Decision trees, Neural Networks, Bayesian Networks, and advanced ensembles of predictive stages like LogitBoost and Adaboost techniques. Among these five techniques, LogitBoost and Adaboost models give true positive rate of 88.8% for predicting patients with suicide ideation. The results of this praxis show that clinicians can use the model as diagnostic tool to screen patients with suicidal thoughts for suicide risk management. The research also shows that economic productivity cost could be reduced by 6.67% with the output of the predictive model and early therapeutic treatment from the healthcare professional.

Author

• Ta, Paul

Language

• en

Date created

• 2022

Type of Work

• Dissertation

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Engineering Management

Degree

• D.Engr.

Advisor

• Dimitriou, Aristos

Committee Member(s)

• Kamp, John
• Dimitriou, Aristos
• Etermadi, Amir

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/tt44pn62t