Application Notes for Intel® oneAPI Math Kernel Library Summary Statistics

ID 772991
Date 12/04/2020
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

Detecting Outliers in Datasets

Datasets may contain outliers or bad observations that do not belong to the distribution to be analyzed. The cause may be an unreliable process of data collection, as in the case of using micro-array technologies for measurement of gene expression levels, or intentional actions, such as network intrusion. Outliers can lead to biased estimates and wrong conclusions about the object.

To process datasets with outliers, you can choose between the BACON outlier detection algorithm and robust methods considered in the following sections.

Product and Performance Information

= = = = = = = = = =

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.

Notice revision #20201201

= = = = = = = = = =