Provides an inside view of a genetic algorithm package from Frederick, MD-based Ward Systems Group.

Within the academic world, efforts continue to unite these two technologies. In a recent book chapter, University of Alabama mathematician James Buckley and Ibaraki University computer scientist Yoichi Hayashi define a fuzzy neural network as a net whose input signals and/or interconnection weights are fuzzy sets. Appearing in Fuzzy Sets, Neural Networks, and Soft Computing (a 1994 Van Nostrand Reinhold publication), their paper outlines the specifications of such a net, shows how one might apply a fuzzy net within the context of industrial control, and lays out a program for further research. Given the current mind-boggling rate of progress, it wouldn't be at all surprising to find a concrete applications of their work within these pages by the time we turn the spotlight of fuzzy logic and neural nets next year.

In the meantime, the articles within this issue's focus provide a sense of real-world excitement in both fields. The Feature Article, "Lesions in Neural Networks," gives you a look at a line of neural net research that may one day prove beneficial to unfortunate individuals who have suffered brain damage. This research might enlighten us about the functioning of the intact brain as well. William Siler's "Fuzzy Reasoning" shows how fuzzy logic dovetails with expert systems. Siler also lists some important applications of the union of these technologies. Gene Dilmore's "Fuzziness in Real Estate" applies fuzzy logic to an everyday area -- property appraisals. Dilmore describes how belief graphs neatly summarize an estimator's thought process. In "Building Large-Scale Fuzzy Systems," Fred Watkins explains issues associated with combinatorial growth in fuzzy logic-based industrial strength systems.

Our other features are also designed to convey the excitement of their subjects. Jeff Rapaport reports on "Rough Sets," a technology often confused with fuzzy logic. Rapaport shows how rough sets-based analysis works with a database of cases and returns a set of rules. Chuck Williams concludes his two-part series on "Smart Front-Office Applications" by examining the technologies required to build leading-edge client/server software. Hal Berghel continues his "Cybernautica" series with a review of the origins of cyberspace.

This issue marks the beginning of new category, "First Looks." Under this banner, we'll bring you in-depth coverage of new Artificial Intelligence-related products as they emerge from beta testing and hit the market. Add this category to our "Vendor's Forums" (which present an inside view of product development), software reviews, Product Updates, and Buyer's Guide, and you have a full spectrum of information on the packages primed to solve tomorrow's problems. In our first "First Look," Lisa Lewinson tells you about GeneHunter, the new Genetic Algorithm tool from neural net powerhouse Ward Systems.

Read and enjoy!

Joseph Schmuller