Good commentary, thanks for sharing! Two quick pieces of advice: (1) Put the phone on a small table next to the board; (2) Don't need to repeat so many times that you are just a beginner :)
@user-nq9tb4zo4e7 жыл бұрын
Thanks for your advice Some people have difficulty to forgive commentator's mistake in this field I want to make sure people understand I make mistakes especially comparing to those pros but thank you,
其實這種類神經網路的 AI 已經不是用演算法了,圍棋這麼複雜的沒有合適的演算法來推演。 類神經網路其實是模擬較低階生物大腦的神經元細胞運作的模式來學習事情。 會用低階的生物大腦的運作,是因為低階生物的腦細胞的運用效率很高。 牠們在很有限的細胞數量去做很多複雜的事情,然而這樣的類神經網路是怎麼辦到學習行為? 是怎麼累積到這麼複雜的計算行為?其實我們人類目前也不知道... 只知道重複的訓練 AI 它可以不斷的學習,然後累積的學習經驗會展現出現在人類遠做不到的事情...