A decision support system for diagnosis of hepatitis disease using an integrated rough set and fish swarm algorithm

Abstract

Healthcare data analysis is a primary concern. It leads to multiple levels of knowledge extraction for decision support systems because of the presence of uncertainties.Therefore, this paper integrates the rough set and artificial fish swarm optimization to develop a decision support system that handles uncertainties present in an information system. In the initial stage, the artificial fish swarm-the rough set procedure is implemented in finding vital features. Further, in the second phase, the rough set uses these vital features to develop a decision support system. The above model is analyzed over hepatitis B disease. The proposed model attains an accuracy of 92.4%. Further, the proposed model is compared with the classical rough set, decision tree, and artificial fish swarm-decision tree model. The accuracy obtained is 88.9%, 83.3%, and 90.8%, respectively. The proposed model has a greater accuracy of 3.5% than the rough set model and has a greater accuracy of 9.1% than the decision tree model. Simultaneously, the proposed model has 1.6% greater accuracy than the artificial fish swarm-decision tree model. Therefore, it is believed that the projected decision support system may be used to prevent and detect hepatitis B diseases.

A New Data Science Model With Supervised Learning and its Application on Pesticide Poisoning Diagnosis in Rural Workers

“…The potential of RST in the multidisciplinary field of Machine Learning, involving the selection of relevant attributes and decision-making, has been demonstrated in [28], [29], [30], [31], [32], [33], [34].…”