Open AccessMODIFIED PROBABILISTIC NEURO-FUZZY NETWORK FOR TEXT DOCUMENTS PROCESSING
Author(s) -
Yevgeniy Bodyanskiy,
Iryna Pliss,
Valentyna Volkova
Publication year - 2014
Publication title -
computing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.184
H-Index - 11
eISSN - 2312-5381
pISSN - 1727-6209
DOI - 10.47839/ijc.11.4.582
Subject(s) - computer science , probabilistic logic , probabilistic neural network , artificial intelligence , artificial neural network , text processing , inference , fuzzy logic , data mining , machine learning , time delay neural network
We consider the problem of text processing (classification problem) using the modification of the probabilistic neural network proposed by D. Specht. Since classes often overlap during texts processing, there were introduced the modification that implements a procedure of fuzzy inference. For this into the network were introduced two additional layers. The results of the outputs of the network are levels of belonging each text document to each of the possible classes.