HMFC : hybrid modlem-fuzzy classifier for liver diseases diagnose

Abstract EN

Diagnose of Common diseases like liver is very complicated because its symptoms can be vague and easily confused with other health problems.

In some cases, a person may have no symptoms at all but the liver may already have suffered significant damage.

Many mathematical approach can deal with vagueness and uncertainty like fuzzy set theory (FS) although FS is good for classification but it requires more knowledge and experience in setting rules of inference based.in this paper an intelligent diagnosing model called hybrid modlem2-fuzzy classifier (HFMC) is proposed to enhance risk classification accuracy of liver diseases.

Rough set approaches (RST) used to generate automated perfect rules to improve accuracy of classification of FS.

The proposed model implemented in two phases, in first phase rules generated by laplace-Modlem2 and use othor RST algorithmes (LEM2, Laplace-Modlem2) as comparaison way .in second phase build fuzzy inference system based on the rules generated in first phase.

The proposed model gives result of 99.14 % classification accuracy of rule generation