Fuzzy logic algorithms contain a vast number of complex calculations, and can be powered by a number of different types of AI processing systems, including neural networks. Although the programming side of the fuzzy equation is murky at best, anyone can understand the basic concepts of how this form of AI works.
Rule-Based Fuzzy Systems
Fuzzy Logic Applications are based on rules, like many other AI programs, with a few important distinctions. As the name infers, a “fuzzy” algorithm does not provide a definite answer to a Yes/No question. Instead, this type of AI algorithm uses rules to infer meaningful results based on observations and other data. In essence, it is codified common sense, since it applies rules in much the same way as humans intuitively apply common sense to everyday problems.
When is Fuzzy Logic Useful?
Fuzzy logic is useful when there is a high volume of data to process, but is most appropriate when the processing is not expected to yield a small number of expected results. For example, a program to sort clothing into colors would not require a fuzzy system , since yellow is always yellow, and blue is always blue. A program to sort clothing by the intended gender of the wearer, however, might require a fuzzy logic algorithm, however, due to the fuzzy nature of the concept.
Most people can determine the gender of the owner of an article of clothing by observing subtleties in the cut and color of the item, but articulating the ‘rules” is a difficult task. To use fuzzy logic, the programmer has to build a large list of these rules into the algorithm, so that the program weighs the options and chooses the most likely result. To evaluate a pair of pants, for example, the program might check the brand name of the garment to see if it is made by a manufacturer that mostly caters to one gender or the other. Other criteria might include the color of the garment, the size of the garment and any accessories and embellishments.
Fuzzy Logic Algorithms Fuzzy systems provide a unique set of computer-based decision skills to users. This is particularly useful in business intelligence, (AI Improves Decision Support Software) which is used to help managers make the best decisions based on trends and other pieces of information.
