An efficient disease prediction and classification using feature reduction based imputation technique

Analysis of a medical dataset having missing values and then filling the missing values through different approaches exists in the literature. However, the classification accuracies achieved using these approaches have not been so promising when analyzed. It is this reason; which implicitly motivated us to study and address new methods for imputation. In this paper, we propose an approach for efficient imputation and classification performing imputation applying class based clustering strategy. We call our method as IMputation Based on Class Based Clustering (IMCBC). The classification accuracies achieved are promising compared to state of art approaches in literature when performed on benchmark datasets.