Fuzzy Logic: A Practical Approach [With *]

Fuzzy logic results from probable classifications. Fuzzy logic is about classification. Words are used to build context. Machines must follow rules that humans find intuitive. For example what is an oval. A person can be shown a few examples of an oval and know instinctive which shapes are oval. Fuzzy rules, such as, convex, continueous, closed, etc, probabilities must pattern match to product an action that confirms the object is oval. Mycin is an expert system with both an inference engine and a knowledge base engine. Mycin uses a fuzzy logic inference engine and derives a best solution through a series of questions and answers. In the 1970's Mycin was used to advise physicians on the treatment of bacterial infections of the blood and meningitis. Mycin uses information about specific symptons and the outcomes of laboratory tests then recommends a certain course of antibotics. The system outputs sentences but the user inputs words. Mycin asks facts about a specific patient. The inference engine has a series of rules and each rule has an action. When a rule is activated the pattern part matches the database. If a pattern match occurs an action results. More questions can result from actions, a backward chaining expert system.

I enjoyed the approach here. The author uses simple, rather concrete examples that allow one to see how fuzzy logic systems work. With the shareware given here it also shows that such systems don't take a lot to construct or work with. As one reviewer says the book doesn't have much on the mathematical theory behind fuzzy logic systems, but the author meants such as a practical approach. And the mathematical theory definitely works as much too much to explain in a work like this. Additionally, the main advantage of using fuzzy systems works as that you don't have to use so much math and bury everything in equations and cryptic symbols. This sort of example makes it easier to see this and work with it. The knowledge builder shows some ideas as to how one can generate if-then rules in a simple fashion. The decision maker makes the program practically. And the fuzzy cognitive map shows you how fuzzy logic systems can apply to scientific modeling easily.