Prediction of Days in Hospital Dengue Fever Patients using K-Nearest Neighbor

Abstract

Dengue fever is found more in tropics and
subtropics area. The World Health Organization (WHO) noted
that Indonesia is as the highest dengue fever cases in South Asian
since 1968 till 2009. The treatment of patient dengue fever in
hospital had spent rest time or the days in hospital in variation
treatment. Days in hospital often becomes questions from the
patients’ family. To predict of days in hospital is maybe to know
the capacity for the long period strategies. In this research, we
used machine learning approach to predict the period of dengue
fever patients. One of the machines of learning method is KNearest Neighbor (K-NN). In this case, we analyze the result of
implementation K-NN method that used to predict period of days
in hospital patients for dengue fever. This paper used the result of
laboratory checking of blood complete with patients as parameter
in predict the period of days in hospital dengue fever. The result
of examination used K-NN to show the accuracy levels reach
65,67% with k optimal is k=13.