Skip to main content

Symon.AI Help Center

Outlier

Abstract

Identify outlier values in your data.

Identify outlier values in your data.

When the Outlier type is set to All, this tool looks at the selected speed and applies a subset of algorithms to set the outlier and probability values.

Note

Probability is the number of outliers that are determined True over the total number of outliers. The outlier value is set to "True" if the probability is over 50%. The probability is the proportion of algorithms that identify that value as an outlier. For example, if there are six algorithms and five out of the six algorithms return TRUE, that the value is an outlier, then the Probability is 83.33% with an Outlier value set to TRUE.

You can optionally use the Analyze by field for more control over how Symon.AI identifiers outliers. When you select a column to analyze by, Symon.AI suggests outliers based on groups instead of the entire data set. 

For example, let's say you have a data set with 3 columns:

  • seller ID

  • pay component

  • country

If you want to find outliers in pay components, it would be useful to analyze this by country, since an outlier in one country may not be an outlier in another.

When to use this tool

Use when you want to find values in your data that might be outliers. You can then investigate why these values are outliers and, optionally, remove them.