AlphaGo Zero原理淺析

AlphaGo Zero 論文：《Mastering the game of Go without human knowledge》 AlphaGo與AlphaGo Zero主要有以下幾點不同： AlphaGo中用了３個policy network,AlphaGo Zero只用了一個reinforcement learning network AlphaGo Zero將policy network

>>阅读原文<<