Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Cheating Detection Research in Chess
Explore using CentiPawn Loss to detect chess cheating. This talk covers board position complexity, player ELO estimation, and variance, with applications for a popular Chrome extension.
Cheating Detection in Chess is an interesting ML problem, and my interest in it began because I was considering a new feature to add to Chess Predict chrome extension as a way to differentiate from competition. My discussion and research slides will showcase my work in detecting chess cheating detection, specifically utilizing what I refer to as a CentiPawn Loss (the difference between the engine score for the best possible move, and the engine score for the players move), with the assumption that the smaller this CPL, the closer to optimal the player is playing.
Chess Predict offers AI-powered chess analysis from game screenshots.
Chess Predict Research employs Closure JavaScript for optimized Google Slides font loading metrics.
Compose Email
Loading recent emails...