Automated Mental Health Assessment (AMHA) to Predict Destructive Behaviors

Project: Research project

Project Details

Description

This proposal has two major goals, each with clear sub-goals. These goals and subgoals are necessary for developing a deployable AMHA (schematized in Figure 1):
(1) Identify relationships between NLP, RPT, and MH/DB variables for AMHA
1.1 Produce population estimates of sensitivity/specificity for RPT prediction of the broad array of DB and MH using standard statistical vs. machine intelligence approaches;
1.2 Identify specific NLP variables from a population sample that match validated experimental tasks, and can be used instead of these tasks to estimate RPT variables;
1.3 Optimize machine/deep learning for prediction of MH/DB when NLP is used for RPT variable computation.
(2) Test the integration and compatibility of AMHA components and their hardening
2.1 Shift from concierge prototypes of hybrid AI into an integrated application programming interface (API);
2.2 Assess the API with dev ops simulations for scaling, updating, continuous use
StatusActive
Effective start/end date2/22/212/21/24

Funding

  • Office of Naval Research (N00014-21-1-2216)

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.