Artificial Neural Net - Backpropagation (C++)

Post those lines of code you feel like sharing or find what you require for your project here; or simply use them as tutorials.
sudi
Posts: 1686
Joined: Fri Aug 26, 2005 8:38 pm

Post by sudi »

I guess i will try to put a neural network in my new game(crazy geo). but i will use a genetic algo to evolve them that way the game will hopefully get harder over time. would be really cool if the enemies try to get at u and doge ur shoots.
We're programmers. Programmers are, in their hearts, architects, and the first thing they want to do when they get to a site is to bulldoze the place flat and build something grand. We're not excited by renovation:tinkering,improving,planting flower beds.
JP
Posts: 4526
Joined: Tue Sep 13, 2005 2:56 pm
Location: UK
Contact:

Post by JP »

Nice one blindside, i've wanted to do some neural network stuff for a while but never got round to it, too busy with other things and didn't really know where to start to be honest!

Apparently games companies are trying to use (at least some form of) neural networks more now (according to Christian and his 'contacts' in Ubisoft) but in the past they've always been deemed too much work to construct and train and then you're not guaranteed to get any really sensible or desired behaviour out of it at the end, especially if you're dealing with something like FPS enemy AI which is pretty complex.

Anyway, would be cool to somehow get this into IrrAI ;)
Image Image Image
BlindSide
Admin
Posts: 2821
Joined: Thu Dec 08, 2005 9:09 am
Location: NZ!

Post by BlindSide »

Yeah would be great if you can shove this into irrAI somehow. Are they really considering neural networks now? It would've been taboo a few years ago. I heard alot of AI programming job applicants get looked down upon simply by showing interest in neural networks.
ShadowMapping for Irrlicht!: Get it here
Need help? Come on the IRC!: #irrlicht on irc://irc.freenode.net
JP
Posts: 4526
Joined: Tue Sep 13, 2005 2:56 pm
Location: UK
Contact:

Post by JP »

Yeah when i had an interview at a small games company in... erm... Oxford or something.. they were asking if i'd ever tried using a neural network for something like racing AI and i said no i've not tried it i always thought it was very difficult road to go down due to all the uncertainties and their AI programmer said that he'd tried to get it working and it worked well in some cases but terribly in others, probably due to over training or something and as it's such a fine line to get the training right with lots of trial and error it's a pretty huge task to try!

I'll keep a copy of your code with the rest of my AI stuff and when i get a chance, probably not for a while, we'll have to think of some way of using this :lol: I guess that sounds like we're just using it for the sake of it... but i think we could find a suitable application for it, maybe something simple like Pong AI would be a good place to start!
Image Image Image
rogerborg
Admin
Posts: 3590
Joined: Mon Oct 09, 2006 9:36 am
Location: Scotland - gonnae no slag aff mah Engleesh
Contact:

Post by rogerborg »

AI Nerd: And the great thing about this neural net is that I don't really know how it works, so it's not deterministic how it will perform or respond in any given situation. *SNORK* [takes hit of inhaler]

QA manager: I. Will. Kill. You.
Please upload candidate patches to the tracker.
Need help now? IRC to #irrlicht on irc.freenode.net
How To Ask Questions The Smart Way
JP
Posts: 4526
Joined: Tue Sep 13, 2005 2:56 pm
Location: UK
Contact:

Post by JP »

xD sooo true!
Image Image Image
Post Reply