There is a vast, repeat vast, literature on computational schemes using neural nets, which are capable of 'being trained' and 'learning' from data that they process.
Laurene Fausett's book, Fausett, Laurene V. Fundamentals of neural networks. Englewood Cliffs, NJ: Prentice-Hall, 1994. is, I think, the best introduction. It is not entirely current. She apparently has a revised edition in the works, and the later edition published by Pearson in India is somewhat more current, but the 1994 edition is more than adequate for getting started. Perhaps worth noting explicitly is that neural networks can be, often are, simulated on z/Architecture machines. The "chip" described was in fact simulated in this way. John Gilmore, Ashland, MA 01721 - USA ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
