projects

dsaa 2020 paper on multiclass classification baselines for anomaly-based network intrusion detection. the model distinguishes legitimate, direct, and obfuscated traffic at up to 95% accuracy.

a notebook of a site. rebuilt on next.js with hand-drawn chrome that re-sketches on every visit.

simplifying the scripting-to-function process. machine learning recognizes the inputs and outputs of scripts and produces a usable, savable function.

sentiment analysis of public opinion on charter schools, using forty years of newspaper articles and python.

sentiment analysis on the twitter stream. parked after the api revoked credentials.

a small program for tracking stocks, written to learn by doing.

scratchpad for libraries and skills. i prefer to learn while building, so this folder keeps growing.