Frequently Asked Questions

Is Feature Scaling Required For the KNN Algorithm

Yes, Feature Scaling is required in KNN Algorithm.

Feature Scaling is a process of normalizing the range of independent variables (features) in the dataset.

In the KNN algorithm, distances like euclidean distance are measured between different points. Now let us consider there are two features in the dataset, one of which ranges from 0 to 1 and the other ranges from -200 to 300. This difference in magnitude will affect the performance of the KNN algorithm. Hence all the features are required to be in the same range which can be done by feature scaling.

Other Popular Questions