Information Mining is one of the preeminent descriptive parts of automatic order and identification. It embroils information mining calculations and methods to inspect restorative information. Lately, liver protests have lopsidedly increased and liver diseases are complimenting one of the most human afflictions in various nations. Early determination of Liver Disorder is basic for the welfare of human culture. This grievance ought to be considered earnestly by setting up canny frameworks for the early analyze and anticipation of Liver infections. The robotized grouping framework endures with absence of exactness results when contrasted and careful biopsy. We propose another model for liver disorder order for breaking down the patient's restorative information utilizing cross breed counterfeit neural system. The therapeutic records are arranged whether there is a plausibility of presence of sickness or not. This proposed strategy utilizes M-PSO for highlight choice of information factors and M-ANN calculation for sickness order. The exhibited crossover approach improves the precision when contrasted with existing order calculations. This paper concentrates on feature selection on using PSO algorithm. © BEIESP.