Training Effectiveness Measurement Software

Subsequent Phase II Proposal, extension of Phase II contract (Navy Contract # N00014-13-C-0138). This effort will use advanced machine learning algorithms to measure training effectiveness of a desktop-based trainer (Virtual Task Trainer-VTT) designed to teach valve maintenance to SWOS trainees. UC will develop software based on dynamic Bayesian networks (DBN) to track skill proficiency of trainees across multiple lessons. We will design experiments to collect data from SWOS trainees on the VTT to train and validate the DBN skill prediction models. UC will leverage its expertise in eye-tracking analysis to improve the fidelity of our assessment of the trainer. Phase II Base will produce an alpha version of the software that will be demonstrated to SWOS SMEs. In Phase II Option 1, the team will work on maturing the software for integration with the VTT and also expand the capability of the software to perform assessment of existing and future SWOS trainers. The effort will empower SWOS to quantify and qualify effectiveness of their trainers and appraise recent expenditure by OPNAV N96 and the Surface and Expeditionary Warfare Training Committee (SEWTC). We envision such software will overall reduce long-term training costs by aiding in the design of better trainers.