Articles → MACHINE LEARNING → Outliers In Data In Machine Learning

Outliers In Data In Machine Learning

This article explains outliers in data.

What Is An Outlier?

Outliers are unusual data points that differ significantly from the majority of the data and do not follow a specific pattern.

Outlier Types

The Outliers can be categorized into the following types.

Univariate outliers
Multivariate outliers
Point (Global) outliers
Contextual (Conditional) outliers
Collective outliers

Univariate Outliers

Univariate outliers are extreme values that significantly differ from other values in a single variable.

For example, in a dataset of exam scores where most scores fall between 40 and 90, a score of 5 would be considered a univariate outlier.

Multivariate Outliers

Multivariate outliers are data points that appear normal when each variable is considered separately, but are unusual when the combination of multiple variables is analysed together.

Point (Global) Outliers

Point (Global) outliers are individual data points that lie far away from the overall distribution of the dataset and differ significantly from most other observations.

Contextual (Conditional) Outliers

Contextual (Conditional) outliers are data points that are unusual only within a specific context or condition, not when considered globally.

For example, A temperature of 30°C is normal in summer but would be a contextual outlier during winter.

Collective Outliers

Collective outliers are a group of data points that are unusual together, even if individual points seem normal.

Posted By -	Karan Gupta

Posted On -	Sunday, January 25, 2026

Query/Feedback

Your Email Id		**

Subject		*

Query/Feedback	Characters remaining 250	**