A0255
Title: A Jensen-Shannon divergence based k-NN algorithm for missing value imputation in compositional data
Authors: Michail Tsagris - University of Crete (Greece) [presenting]
Abstract: A novel non-parametric method to impute missing values or rounded zeros in compositional data is suggested. The method is based on the k-NN algorithm, utilizes the Jensen-Shannon divergence, and employs the Frechet mean to allow for more flexibility in the estimation process. As an extra feature, the hyperparameter k can be self-adaptive depending on the pattern of missing values or rounded zeros. Unlike restrictive parametric models, the proposed method makes no assumption about the structure of the data and, most importantly, it is applicable even when compositional data contain structural zero values. Through simulation studies using artificial and real data, the proposed algorithm is superior compared to two competing algorithms at various settings, not only in terms of accuracy but also in terms of computational efficiency.
