Big data clustering: Data preprocessing, variable selection, and dimension reduction