Applying Machine Learning to Predict Midshipmen Performance

Midshipman Researcher(s): 1/C Thomas Maly

Adviser(s): Professor Randy Broussard

Poster #102

Predicting Midshipman performance help USNA identify students likely to succeed in a given environment and scope those who may need further assistance. Machine learning classifiers can improve their accuracy through experience. This project will design an artificial neural network and a Bayesian classifier to predict Midshipman performance at the Naval Academy. It uses inputs provided by the Naval Academy, including High School Data and Academic, Physical, and Military Records, to predict a various things like whether the Midshipman graduates on time or receives their desired service selection. The measured characteristic is the accuracy of the machine learning classifiers.

Full Size Robotics and Controls Engineering #102