Case Based Reasoning untuk Diagnosis Penyakit Gizi Buruk pada Balita

Abstract

This research is conducted to build a diagnose system malnutrition among children under five years old. The system was developed with Case Based Reasoning (CBR). CBR is a case based reasoning system, using old knowledge to solve new problems. CBR can provide new solutions to problems by looking at most similarity case to the previous cases that have been stored in the base case. CBR in this research using a bayesian model indexing to find the type of disease malnutrition among children under five years old, the process of indexing is done to speed up the retrieval process. The nearest neighbor methode used in the process to determine the most similar of cases between new cases and the old cases that have been stored in the database as a case base to be used tratment solution.Tests carried out by using 70 case based were recorded in case of data based and 20 case based serve as a new case. Testing is done with five threshold values. The first scenario is to use threshold ≥ 0.95 system able to produce accuracy 20%. The second scenario is to use threshold ≥ 0.90 system able to produce accuracy 45%. The third scenario is to use threshold ≥ 0.85 system able to produce accuracy 60%. The fourt scenario is to use threshold ≥ 0.80 system able to produce accuracy 75%. The fifth scenario is to use threshold ≥ 0.75 system able to produce accuracy 85%.