Real/Binary-Like Coded Genetic Algorithm to Automatically Generate Fuzzy Knowledge Bases

This paper presents the results of the implementaion of a combination of a real-coded and binary-like coded genetic algorithm (RBLGA) to automatically generate fuzzy knowledge bases (FKB) from a set of numerical data. The algorithm allows one to fulfil a contradictory paradigm in term of FKB precision and simplicity (high precision generally translates into high complexity level) considering a randomly generated population of potential FKBs. The RBLGA is divided in two principal coding ways: 1) a real coded genetic algorithm (RCGA) that maps the fuzzy sets repartition and number (which drives the number of fuzzy rules) into a set of real numbers and. 2) a binary like aenetic algorithm that deals with the fuzzy rule base (a set of integer numbers). The RBLGA uses three reproduction mechanisms, a BLX-α, a simple crossover and a fuzzy set reducer. The RBLGA is validated through a theoretical surface and, funally, applied to a set of experimental data.