Machine Learning with Scikit-Learn - [Part 38]

In this tutorial we're going to delve into another preprocessing method that we're going to implement on our dataset with scikit-learn. This method is called normalization.

Normalization is applied on the values of feature vectors to bring them on a common scale; and this makes the training process much more efficient and less prone to error. We're talking about two most used normalization methods:

• L1 - or the Least Absolute Deviations
• L2 - or the Least Squares

First, we're taking a look on our data. Then we apply each of them on our data. L1 and L2 are available with the preprocessing module in scikit-learn. Ultimately, we look at the output of the code, comparing the results for these two methods.

We verify the methods by taking an example from each output and adding them together. And the result should be equal to 1 in both cases. Please watch the video for the full tutorial. A link to the code is also available.

---

To stay in touch with me, follow @cristi

---

Cristi Vlad Self-Experimenter and Author