Game-AI Forum

GomokuNet: A Novel UNet-style Network for Gomoku Zero Learning via Exploiting Positional Information and Multiscale Features Since the tremendous success of the AlphaGo family (AlphaGo, AlphaGo Zero, and Alpha Zero), zero learning has become the baseline method for many board games, such as Gomoku ...

https://medium.com/@ameyapimpley98/jinxings-interview-60d12f67277a This is an interview conducted by me with Jinxing, who is the author of the Go program Golaxy. Golaxy is considered to be one of the strongest available Go engines in the world and has won many championship titles such as UEC and Fuj...

The 2022 International Computer Games Association events are fast approaching: The *2022 Computer Olympiad *(online, July 23-29). The Call for Participation is here <https://icga.org/?page_id=3468>. Registration is now open (until July 14). The *World Computer Chess Championships* (Vienna, July 23-...

Interesting paper: https://www.deepmind.com/publications/policy-improvement-by-planning-with-gumbel AlphaZero is a powerful reinforcement learning algorithm based on approximate policy iteration and tree search. However, AlphaZero can fail to improve its policy network, if not visiting all actions a...

flok wrote: ↑

Fri Apr 22, 2022 8:20 pm

Ok. But which parent should I use?

The parent from which you are selecting a move.

Hi, Using a transposition table is a good idea, and is rather common. It should make the program stronger. You should not have to keep pointers to parents. If I understand what you do correctly, you may get very different "parent_visit_count" for every child of a node. This does not seem good. Also ...

Thanks for sharing the news. I plan to participate.

Thanks Yamashita san for reporting the results.

And let's implement reconnection support for next year.

I am looking forward to the next edition. I hope I'll be able to travel to Japan then.