Archetype analysis: A new subspace outlier detection approach

The problem of detecting outliers in multivariate data sets with continuous numerical features is addressed by a new method. This method combines projections into relevant subspaces archetype analysis nearest neighbor algorithm, through an appropriate ensemble the results. Our able to detect anomaly simple set linear correlation two features, while other methods fail recognize that anomaly. performs among top extensive comparison 23 state-of-the-art outlier detection algorithms several benchmark sets. Finally, novel industrial introduced, and carried out improve fit footwear, since this kind has never been fully exploited anthropometric field. • Archetype combined proximity-based techniques for detection. Experimental results large show good performance Outlier foot measurements improving footwear.