K-Means Clustering Using Approximate Addition