Outlier Mining in Large High-Dimensional Data Sets