KNN Method for Handling Right-censored Data
Abdul Kudus (a*), Suliadi (a), Yayat Karyana (a), Caecilia A Rahman (a)

a) Dept. of Statistics, Universitas Islam Bandung, Jl. Ranggagading No. 8 Bandung 40116, Indonesia
*abdul.kudus[at]unisba.ac.id

Abstract

The K-Nearest Neighborhood (KNN) method can be used to impute censored data. KNN imputations are designed to find the nearest neighbor^s K from incomplete data to the entire occurrence of a data, then fill in the missing data with the events that are most similar to their neighbors, if the target variable (or attribute) is categorical then the imputation refers to the majority of neighbors but if the variable is numeric then the imputation uses the average of the nearest neighbors. This article discussed the use of KNN imputations in the survival time of right-censored patients based on the averages of as many K as nearest neighbors. The distance from censored data to its neighbor was calculated based on two independent variables. The smaller the distance, the data becomes the closest neighbor because it has similar characteristics. The imputation value for the censored data was just the average of its neighbors which is forty six neighbors around the censored data.

Keywords: Censored data- Imputation- K-Nearest Neighbour

Topic: Mathematics