Elaborate Sklearn Pipelines

Code of this talk is on Github

<aside> 💡 There are so many steps and decisions to take when training an ML model. → Get some inspirations here.

</aside>

Pieces of a Data Science project

Common preprocesing code

Let’s say we want to do step (6). We can make a function to do this job, but in in this, and many other cases, we have to memorize and apply it to the test set → explanation here

You can build your own transformer objects → explanation