Neural Networks

Started by
3 comments, last by Kylotan 14 years, 5 months ago
hi guys, i have been doing some research into neural networks, and i'm not quite sure how to set one up. Anyone know where i can find the basics and a bit of code? jamescobras
Advertisement
Well, Wikipedia is always a great place to start.
Plenty of info and links.
Gotta love Google, too.

Dave Mark - President and Lead Designer of Intrinsic Algorithm LLC
Professional consultant on game AI, mathematical modeling, simulation modeling
Co-founder and 10 year advisor of the GDC AI Summit
Author of the book, Behavioral Mathematics for Game AI
Blogs I write:
IA News - What's happening at IA | IA on AI - AI news and notes | Post-Play'em - Observations on AI of games I play

"Reducing the world to mathematical equations!"

Self Organising Feature Maps I found to be the simplest to get my head around:
http://en.wikipedia.org/wiki/Self-organizing_map
I would have thought a simple 2-in, 1-out perceptron would be even simpler still. You can implement one of those in about 20 lines of C.
- 2 arrays, one for input values and one for weights, representing the input layer.
- A routine that calculates the output by multiplying all inputs by weights and adding them up.
- A threshold routine that takes the previous output and generates a 1 or a 0 based on the value.
- a perceptron learning routine to adjust the weights a certain amount when the output is wrong, based on the training data
- a few lines of glue code to set up the initial values and then call these things in the right order for a given problem (eg. simulating an AND or OR gate)

This topic is closed to new replies.

Advertisement