Implementation of Modified K-Nearest Neighbor for Diagnosis of Liver Patients

Sari

Number of patients with liver disease in the world
is very high. In the early stages, liver disease is difficult to
detect. Early diagnosis of the liver disease may help in
preventing and treating sufferers. To diagnose liver disease
can be done with a blood test. Based on data from this
analysis, the results can assist in determining patients with
liver disease. This study uses data Indian Liver Patient Dataset
(ILPD) taken from the UCI Machine Learning Repository. We
used Modified k-Nearest Neighbor to classify into two classes,
namely sufferers and non-sufferers. The amounts of data used
in this study were 583 records. Tests performed by dividing
the training data and test data to 50:50, 60:40, 70:30 and
80:20. Results of tests performed can classify with a good
degree of accuracy reached 85.14% with a ratio of 70:30 and k
= 3.