A grey-based nearest neighbor approach for predicting missing attribute values

Abstract

In this paper, we propose a grey-based nearest neighbor approach to predict missing attribute values in an accurate manner. First, the nearest neighbors of an instance with missing attribute values are found through grey relational analysis. Accordingly, the known attribute values derived from these nearest neighbors are chosen to infer those missing. The Iris flower dataset was used to demonstrate the performance of the proposed approach. Experimental results show that our method performs better than both multiple imputation and mean substitution.