Hi, I'm currently writing a literature review for my Doctorate thesis and was wondering what books people here would recommend for information on the history and more recent development of ANN's in all their shapes and sizes? I have a basic working knowledge of ANN's and have implemented a few, as my undergraduate thesis involved working with ANN's and different training techniques.

If you're already familiar with NN, and are looking for some heavy duty for thorough book, then Bishop's "Neural Networks for Pattern Recognition" is your best bet. It's intense on math though, but if you're at Ph.D. level you should arguably be able to cope.

I was going to recommend the same book as alexjc also. It's a very good book.

You may also want to look around sites like that of Cambridge University Press, see here for example:

http://www.cambridge.org/uk/catalogue/searchResult.asp?hl=en&q=&sort=Y&criteria=Neural+Network&bntG.x=0&bntG.y=0&searchType=onthissite

Or just www.cambridge.org if you want the bookshop front.

They tend to have some really cutting edge publications there, so for a doctorate it's really the sort of place you should probably be looking.

Thanks for the helpful suggestions guys :)

I think that real progress on neural networks has slowed in recent years. I suggest looking into related areas instead. A good example is model ensembles. See, for instance: Combining Models to Improve Classifier Accuracy and Robustness, by Dean W. Abbott. Other areas of interest might be: dealing with missing values and attribute selection.

Quote: Original post by Snikks
I'm currently writing a literature review for my Doctorate thesis and was wondering what books people here would recommend for information on the history and more recent development of ANN's in all their shapes and sizes?

I have a basic working knowledge of ANN's and have implemented a few, as my undergraduate thesis involved working with ANN's and different training techniques.

I just finished a PHD and my dissertation project involved machine learning with ANN to generate procedural game content suited to user preferences. I'm assuming you want ANN stuff geared specifically toward games?

This page on AIWisdon.com list major books that are either wholly or have chapters devoted to ANNs in mainstream type games:

http://www.aiwisdom.com/ai_neural.html

^ Those will be quite useful because they cover a lot of the practical implementation issues. If you want some good references and background material, check the following sections from this journal paper: "Machine Learning in Existing Games" and "Procedural Content in Existing Games":

http://eplex.cs.ucf.edu/publications/2009/hastings.ieeetciaig09.html

BTW, do you have any ideas on a dissertation project? The main research area for ANNs in games is adaptive AI, but more recently intelligent procedural content generation techniques are beginning to catch on. Or will you do something completely different than those? I'd be interested to hear about it. =)