This project was worked on as part of Stanford’s CS 229: Applied Machine Learning, led by Professor Andrew Ng. Abstract, full report, and code repo presented below.

Abstract

This project aims to build a model for predicting the risk of illegal substance use in adults. We attempt to use demographic data, as well as personality trait information, to estimate an individual’s likelihood of having made use of illegal substances within the past calendar year. We formulate the problem as a Binary Classification task, and experiment with different approaches, specifically Logistic Regression and Neural Networks, in order to compare the quality of the results, which is relatively high in both instances. Finally, we provide insight into how the results might be improved.

Project Report: Final Project Report

Github link: Predicting the Risk of Illegal Substance Use in Adults