Description of learning space based on interval fuzzy sets and fuzzy inductive learning algorithm ILA

In this paper, based on interval fuzzy sets, the inductive learning space is fuzzified, the distinction interval is newly defined to measure the definability of concepts to be learned. By applying the fining operation of partitions for the basic event space, the value of the distinction interval and the efficiency of inductive learning for concepts to be learned are improved. Based on these, fuzzy Inductive Learning Algorithm ILA is presented for fuzzy inductive learning problems. In terms of the example sets of concepts given by experts, algorithm ILA can learn fuzzy inference rules about concepts to be learned. Finally, the complexity of algorithm ILA is also discussed. (C) 1999 Elsevier Science B.V. All rights reserved.