Newest 'self-play' Questions

0

votes

0answers

13 views

Playing Connect Four with reinforcement learning

I'm trying to do self-play reinforcement learning on a board game called Connect Four and I'm not getting good results so would appreciate some ideas on how to improve. I'm using the PPO algorithm ...

asked Jul 20 at 9:30

BessieTheCookie

101

0

votes

0answers

14 views

In terms of self play which is a harder problem: chess or kung-fu

Giraffe (or Alpha Zero) do a very solid job on superhuman chess. How does the complexity of the "game", especially for simulation in 3d competitive play environments, differ between chess ...

computational-complexity self-play state-space-complexity

asked Jul 1 at 15:41

EngrStudent

1515 bronze badges

2

votes

2answers

241 views

How to fight with unstability in self play?

I'm working on a neural network that plays some board games like reversi or tic-tac-toe (zero-sum games, two players). I'm trying to have one network topology for all the games - I specifically don't ...

reinforcement-learning convolutional-neural-networks deep-rl epsilon-greedy-policy self-play

asked Apr 10 at 22:13

Maras

1315 bronze badges

3

votes

1answer

89 views

How does MuZero learn to play well for both sides of a two-player game?

I'm coding my own version of MuZero. However, I don't understand how it supposed to learn to play well for both players in a two-player game. Take Go for example. If I use a single MCTS to generate an ...

reinforcement-learning self-play muzero

asked Oct 31 '20 at 17:47

Ziofil

1505 bronze badges

1

vote

0answers

44 views

What does self-play in reinforcement learning lead to?

Suppose, instead of playing against a random opponent, the reinforcement learning algorithm described above played against itself, with both sides learning. What do you think would happen in this case?...

reinforcement-learning self-play

asked Aug 2 '20 at 13:16

epsilon-emperor

8901 silver badge13 bronze badges

1

vote

2answers

72 views

Generalising performance of Q-learning agent through self-play in a two-player game (MCTS?)

I'm using Q-learning (off-policy TD-control as specified in Sutton's book on pg 131) to train an agent to play connect four. My goal is to create a strong player (superhuman performance?) purely by ...

reinforcement-learning q-learning self-play

asked Jun 14 '20 at 15:29

Toekan

132 bronze badges

3

votes

0answers

160 views

How to correctly implement self-play with DQN?

I have an environment where an agent faces an equal opponent, and while I've achieved OK performance implementing DQN and treating the opponent as a part of the environment, I think performance would ...

deep-learning reinforcement-learning dqn deep-rl self-play

asked Mar 17 '20 at 12:49

Pell000

311 bronze badge

1

vote

0answers

48 views

Is it possible to use just one policy in a self-play setting? [closed]

I would like to ask is it possible to train an agent under self-playing setting but with just one policy to be trained? What are the foreseeable problems with such an implementation? My concern is as ...

reinforcement-learning self-play

asked Feb 10 '20 at 7:56

Zzy1130

1234 bronze badges

3

votes

0answers

52 views

How exactly does self-play work, and how does it relate to MCTS?

I am working towards using RL to create an AI for a two-player, hidden-information, a turn-based board game. I have just finished David Silver's RL course and Denny Britz's coding exercises, and so am ...

reinforcement-learning definitions monte-carlo-tree-search pomdp self-play

asked Feb 9 '20 at 13:05

Alienator

312 bronze badges

4

votes

1answer

106 views

In AlphaZero, which policy is saved in the dataset, and how is the move chosen?

I've been doing some research on the principles behind AlphaZero. Especially this cheat sheet (1) and this implementation (2) (in Connect 4) were very useful. Yet, I still have two important questions:...

reinforcement-learning alphazero self-play

asked Jan 10 '20 at 23:04

Jonas De Schouwer

1412 bronze badges

3

votes

0answers

43 views

How to deal with nonstationary rewards in asymmetric self-play reinforcement learning?

Suppose we're training two agents to play an asymmetric game from scratch using self play (like Zerg vs. Protoss in Starcraft). During training one of the agents can become stronger (discover a good ...

reinforcement-learning deep-rl self-play

asked Dec 24 '19 at 19:26

Dzugaru

1311 bronze badge

2

votes

1answer

141 views

What will Q-values look like in self-play tic-tac-toe?

This corresponds to Exercise 1.1 of RLBook, and a discussion followed from here. Considering two reward schemes- Win = +1, Draw = 0, Loss = -1 Win = +1, Draw or Loss = 0 Can we say something about ...

deep-learning reinforcement-learning q-learning combinatorial-games self-play

asked Jan 21 '19 at 18:59

pg2455

1893 bronze badges

2

votes

1answer

299 views

Is Python and a single computer sufficient to implement and train an RL-based agent for a turn-based war game? [closed]

I have a steady hex-map and turn-based wargame featuring WWII carrier battles. I would like to improve the given AI part of the game using reinforcement learning. I have a bunch of noob questions. Is ...

reinforcement-learning python game-ai hardware self-play

asked Dec 23 '18 at 7:09

Carrier Battles

1296 bronze badges

1

vote

1answer

108 views

How can I oppose two AI agents with keras / tensoflow?

I am trying to use tensorflow / keras to play a text based game. The game opposes two players that play by answering questions by choosing an answer among the proposed ones. Game resembles this: ...

tensorflow keras self-play

asked Dec 14 '18 at 16:30

Matthieu Raynaud de Fitte

1135 bronze badges

4

votes

2answers

2k views

Why does self-playing TicTacToe not become perfect?

I trained a DQN that learns TicTacToe by playing against itself with a reward of -1/0/+1 for a loss/draw/win. Every 500 episodes I test the progress by letting it play some episodes (also 500) against ...

reinforcement-learning self-play

asked Jun 5 '18 at 21:16

Axel

715 bronze badges

3

votes

2answers

613 views

How can both agents know the terminal reward in self-play reinforcement learning?

There seems to be a major difference in how the terminal reward is received/handled in self-play RL vs "normal" RL, which confuses me. I implemented TicTacToe the normal way, where a single ...

reinforcement-learning game-ai self-play tic-tac-toe

asked May 30 '18 at 10:50

Axel

715 bronze badges

Stack Exchange Network

current community

your communities

more stack exchange communities

Questions tagged [self-play]

Playing Connect Four with reinforcement learning

In terms of self play which is a harder problem: chess or kung-fu

How to fight with unstability in self play?

How does MuZero learn to play well for both sides of a two-player game?

What does self-play in reinforcement learning lead to?

Generalising performance of Q-learning agent through self-play in a two-player game (MCTS?)

How to correctly implement self-play with DQN?

Is it possible to use just one policy in a self-play setting? [closed]

How exactly does self-play work, and how does it relate to MCTS?

In AlphaZero, which policy is saved in the dataset, and how is the move chosen?

How to deal with nonstationary rewards in asymmetric self-play reinforcement learning?

What will Q-values look like in self-play tic-tac-toe?

Is Python and a single computer sufficient to implement and train an RL-based agent for a turn-based war game? [closed]

How can I oppose two AI agents with keras / tensoflow?

Why does self-playing TicTacToe not become perfect?

How can both agents know the terminal reward in self-play reinforcement learning?

Hot Network Questions

Questions tagged [self-play]

Related Tags