Case Based Reasoning untuk Diagnosis Penyakit Ikan Kakap Putih

Abstract

Case Based Reasoning (CBR) is one reasoning from an expert system, namely by reasoning from previous cases that have been stored on a case base to find out the solution of a new case. In case based reasoning there is a retrive process, in the retrieve process there is a similarity process, and to speed up the retrieve process it can use the indexing method. In this research will use the indexing method with Bayesian models and similarity processes using the nearest neighbor method. System testing techniques from this study with two testing techniques namely: the first testing technique using the Bayesian indexing model, the results of the indexing have produced white snapper disease, then proceed with similarity method with the nearest neighbor method used to determine the right solution from the previous case. has been saved on a case base. The second testing technique is without using indexing, the process is only by the nearest neighbor similarity method, the results of similarity in the form of disease and treatment solutions from previous cases that have been stored on a case base. System accuracy for testing with Bayesian model indexing and nearest neighbor similarity with threshold 0,70 is 86% and testing without indexing with Bayesian model with threshold 0,70 is 100%.