What is an outlier and inlier?

A data point that differs considerably from other observations is referred to as an outlier. Outliers can be harmful to machine learning since they can reduce a model’s accuracy, depending on the reason of the outlier. It’s critical to eliminate outliers from the dataset if they’re produced by a measurement error.

An inlier is a data observation that is uncommon or incorrect in comparison to the rest of the dataset. Because it is a part of the dataset, it is usually more difficult to spot than an outlier, and therefore necessitates the use of external data to do so. If any inliers are found, you may simply delete them from the dataset to deal with them.