Some contributions to k-means clustering problems