An Intelligent Hybrid Model for the Construction of Expert Systems in Malware Detection
Paulo Vitor de Campos Souza, Augusto Júnio Guimarães, Thiago Silva Rezende, Vanessa Souza Araújo, Luiz Antônio Ferreira do Nascimento, Lucas Oliveira Batista · 2020
This work has as main objective the development of a hybrid expert system in the identification of malware using the artificial intelligence and fuzzy system approach. Such attacks can harm businesses and individuals by allowing essential data to be misrepresented. To perform the fuzzy rules extraction, the hybrid model was subjected to malware detection tests that were made available in some public datasets and to verify the efficiency of the hybrid approach; it was compared in binary classification tests with hybrid models of fuzzy neural networks and with artificial neural network models. The results obtained in the simulations confirm that the fuzzy neural network approach is feasible to treat malware detection and allows the production of fuzzy rules that can assist in the construction of expert systems.