.com.unity Forums

.com.unity Forums (http://forum.shrapnelgames.com/index.php)
-   Space Empires: IV & V (http://forum.shrapnelgames.com/forumdisplay.php?f=20)
-   -   OT: I made a Tic-Tac-Toe learning AI. (http://forum.shrapnelgames.com/showthread.php?t=21742)

narf poit chez BOOM November 19th, 2004 10:41 PM

OT: I made a Tic-Tac-Toe learning AI.
 
Well, I've never really done an AI before, much less a learning one. http://forum.shrapnelgames.com/images/smilies/smile.gif

It's on my webpage. http://forum.shrapnelgames.com/images/smilies/smile.gif

Didn't take long, once I figured out what I was doing.

Arkcon November 20th, 2004 12:04 AM

Re: OT: I made a Tic-Tac-Toe learning AI.
 
Great. Now expand it to be a checkers learning program, just add a few more states and rules. After that, a chess program is just a few more steps away. http://forum.shrapnelgames.com/images/smilies/wink.gif

narf poit chez BOOM November 20th, 2004 12:11 AM

Re: OT: I made a Tic-Tac-Toe learning AI.
 
Chess AI? Chess AI?! *Blink Blink* *Faint*

It's not a true learning program, all it does is record the part of the board that is three X's or O's in a row when it wins, then next game it checks that list to see if anything on the board looks close to one of those patterns. If it does, it tries to fill in the missing parts. In v2 each AI should also check their enemy's tokens to see if they are close to any winning solution it knows, then if they are, place a peice in that pattern to interfere. Very simple. Just learns patterns and runs them through if-then statements. Have to re-write it for checkers...And I still have a chat program to finish.

Makinus November 20th, 2004 09:43 AM

Re: OT: I made a Tic-Tac-Toe learning AI.
 
try to make a bigger grid (like 4x4 or 5x5) still using 3 tokens to win so it can learn more complicated strategies...

douglas November 20th, 2004 01:21 PM

Re: OT: I made a Tic-Tac-Toe learning AI.
 
That won't work. With a larger grid and still only 3 in a row needed to win, the first player to move can easily win every time.

Kamog November 20th, 2004 11:45 PM

Re: OT: I made a Tic-Tac-Toe learning AI.
 
If it's 4 in a row needed to win, it should work, right? http://forum.shrapnelgames.com/images/smilies/smile.gif

narf poit chez BOOM November 22nd, 2004 05:42 AM

Re: OT: I made a Tic-Tac-Toe learning AI.
 
I could increase the grid and row size, but there really wouldn't be any point. This was just made so I could so if I could do a very basic AI. I did. It's not likely to be expanded; it's more likely I'll make a different AI test program.


All times are GMT -4. The time now is 09:46 PM.

Powered by vBulletin® Version 3.8.1
Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
Copyright ©1999 - 2025, Shrapnel Games, Inc. - All Rights Reserved.