3- Data Scaling 101: StandardScaler vs MinMaxScaler in Machine Learning

posted on 28 Nov 2024

Imagine trying to compare an IELTS score of 7 with a TOEFL score of 50. Without understanding that these exams have completely different scaling systems, any comparison would be meaningless. The same principle applies in Machine Learning: if your features are on different scales, your model can’t make fair comparisons, and your results will suffer.

In this article, we’ll explore two popular scaling methods StandardScaler and MinMaxScaler to understand their philosophy, their mathematics and when to use each one. By the end, you’ll see how these simple tools can make a big difference in your projects.

To read the full version of this article, which has published on Towards AI publication, use the link below:

Data Scaling 101 on Medium