TensorFlow.js implementation of DeepMind's AlphaGo-Zero algorithm for chess

ai-games

(Adavidoaiei Dumitru-Cornel) #1

Este un AI care joaca sah, foloseste neural networks:
Live demo at: https://frpays.github.io/lc0-js/engine.html
Open source code at: https://github.com/frpays/lc0-js/

AI in Games e un subiect interesant destul de vechi, printre tehnicile moderne reinforcement learning, Q Learning, cele mai vechi foloseau tot felul de algoritmi de la grafuri ca si A*.

Un mic AI care joaca X si 0:

Andrej Karpathy are o gramada de articole, demo-uri pe tehnici de reinforcement learning https://cs.stanford.edu/people/karpathy/

Proiect GitHub:

Site cu demo-uri:
https://cs.stanford.edu/people/karpathy/reinforcejs/