Machine Learning NCAA Tournament Predictor ; Tech Companies' Response to Compulsive Device Use
White, Thomas, School of Engineering and Applied Science, University of Virginia
Norton, Peter, EN-Engineering and Society, University of Virginia
Zhang, Miaomiao, EN-Elec/Computer Engr Dept, University of Virginia
Ji, Yangfeng, EN-Comp Science Dept, University of Virginia
Substantial and growing fractions of personal work, play, and entertainment are mediated through digital devices. This development not only affords new sources of information, communication, and amusement, but also new means of interacting with them. Yet with such advantageous developments come new risks, such as compulsive device use.
A perfect bracket in the NCAA Men’s March Madness is a near impossible feat, but machine learning models may yield better predictions. Google Cloud’s annual competition has stimulated improvements in model accuracy, but few people have the means to develop their own machine learning models. The research team designed a prototype website on which users may develop their own machine learning models to predict the NCAA tournament results, selecting training data such as turnovers per game, seeding, or point differentials. The website evaluates users’ models on the bases of historical performance and of team advancement probability. Users can apply these evaluations to improve their models.
To protect their reputations, tech companies respond to allegations that their products induce compulsive device use, which has been correlated with depression and anxiety. Device usage has been correlated with rising depression rates and other mental health effects. Tech companies have introduced device features that ostensibly mitigate compulsive use, but have neither made their products less addictive nor offered users the information they need to take best advantage of these features.
BS (Bachelor of Science)
NCAA Tournament, Device Usage, Machine Learning, Tech Companies, Response to Critics
School of Engineering and Applied Science
Bachelor of Science in Computer Science
Technical Advisor: Miaomiao Zhang
STS Advisor: Peter Norton