Outlier Analysis

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save

Springer+ Basic €32.70 /Month

Get 10 units per month
Download Article/Chapter or eBook
1 Unit = 1 Article or 1 Chapter
Cancel anytime

Buy Now

Price includes VAT (France)

eBook EUR 42.79 Price includes VAT (France)

Softcover Book EUR 52.74 Price includes VAT (France)

Hardcover Book EUR 73.84 Price includes VAT (France)

Tax calculation will be finalised at checkout

Purchases are for personal use only

Notes

Note that higher \(k\) -nearest neighbor distances indicate greater outlierness.

We say “almost,” because the very last distance computation for \(\overline \) may bring \(V(\overline )\) below \(L\) . This scenario is unusual, but might occasionally occur.

Most descriptions in the literature omit the first phase of sampling, which is very important for efficiency maximization. A number of implementations in time-series analysis [306] do order the data points more carefully but not with sampling.

Author information

Authors and Affiliations

IBM T.J. Watson Research Center, Yorktown Heights, New York, USA Charu C. Aggarwal

Charu C. Aggarwal