In this project we attempted to implement a no-calibration face-controlled experimental tool to deliver contingent stimuli and automated participant feedback to very young children.

I was a collaborator on laboratory experiments to identify key performance and behavioral metrics to enhance robot decision making when deployed within human teams. This study takes fairness considerations into account and focuses on optimal resource allocation decisions.

Designed and implemented a compiler for the /Xi_language. The compiler was complete with a lexer, parser and evaluator. Implemented optimization passes to allocate registers, eliminate dead code, fold constants and eliminate common subexpression using Java visitors. Optimizations sped up compiled assembly code by ~40%.

Designed a variant of the /JoCalf programming language with automatic dynamic function memoization features and built an interpreter for it in Ocaml.

Used US county 2016 and 2020 demographics data to predict voter patterns with 80% accuracy. Extracted and analyzed features, tuned SVMs, Neural networks and Decision trees. Then wrote a report on my design process.