Python Machine Learning Tutorial | Centering And Scaling | Databytes

This Python tutorial for beginners will quickly walk you through two data pre-processing techniques for machine learning: Centering and Scaling. When applying machine learning to extract predictions from datasets, the most commonly used pre-processing techniques are centering and scaling. Centering and scaling are essential for model types that assume each feature comes from a standard normal distribution. The topics covered in this video are:

00:00 - 00:58 What is data preprocessing?
00:59 - 01:37 What are centering and scaling?
01:38 - 03:33 When should you use centering and scaling?
03:34 - 08:13 Importing the diamonds dataset
08:14 - 09:34 Running a KNN-model without scaling
09:35 - 14:05 Running a KNN model with scaling

[Try it yourself!]

Check out some interesting reads:

[raimohaikari]

Would it be better to first fit the scaler with train set and after that scale both training and test set with that scaler...