A talk introducing the audience to Scikit-learn as a library aimed at people who know Python at a beginner/intermediate level but are new to machine learning concepts. The goal of the talk is for the audience to leave with an understanding of the foundations of machine learning while respecting how easy it is to make a wrong choice that invalidates your model.
It will be a short background on scikit-learn followed by a livecoding demo where I demonstrate how scikit-learn works and detail common pitfalls.
I will demonstrate ways of coping with problems such as data leakage, the importance of train-test splits, choosing metrics wisely, and explain how cross-validation works and why we use it.
Feedback form: https://python.it/feedback-1596
in __on Saturday 4 May at 10:45 **See schedule**