Datasets often contain anomalies or outliers whose properties are different from those of the regular data points. Such outliers if not handled could lead to bad models which do not work even for the typical non-outlier points. How do we identify outliers in data?
There are many techniques to identify outliers. the Z-score is one of the simplest and one of the most popular techniques for outlier detection that works well for several usecases. This video talks about Z-Score, where it is used, where it does not work and how it can be implemented with simple python code.