Jump to content

  • Log In with Google      Sign In   
  • Create Account

Interested in a FREE copy of HTML5 game maker Construct 2?

We'll be giving away three Personal Edition licences in next Tuesday's GDNet Direct email newsletter!

Sign up from the right-hand sidebar on our homepage and read Tuesday's newsletter for details!


We're also offering banner ads on our site from just $5! 1. Details HERE. 2. GDNet+ Subscriptions HERE. 3. Ad upload HERE.


Neural Net Code


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
No replies to this topic

#1 NickGA   Members   -  Reputation: 122

Like
Likes
Like

Posted 26 April 2000 - 06:54 PM

I am a beginner programmer in Windows programming in C with Direct X. My first game is in the game contest called The Tailgater. The AI needs more work on that game (that''s for sure) For my next project, I want to incorporate some AI (neural nets, fuzzy logic, fuzzy neural net, genetic algorithms, etc.). I''m not sure yet what kind of game it will be, so I am asking you for ideas if you have any. I wonder if a game like mortal combat would be appropriate, or perhaps a space adventure game with traders that learn how to trade better as the game goes along. Of course, anything I make will be free gratis (i''m not a professional game programmer anyhoo and I borrow a lot of free music off the net). I have a neural net structure in C I wrote and you are welcome to it if you like this sort of thing as I do. It is a simple feedforward, backprop network (fully connected). You can easily establish the number of input, output, and hidden nodes. Also, it wouldn''t be difficult to alter the code to make it non-fully connected. I used the sigmodal activation function, but you can change that also if you wanted. Just email me or I''ll check the posts and possibly email you the code if your interested. Perhaps, it''s too slow of a learner (you can specify the learning rate though). Perhaps, it would be better to train it offline first and then train from there as the game progresses... Free gratis of course. Adios Amigos

Sponsor:



Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS