Скачать или смотреть Hyperparameter Tuning in Machine Learning | Grid Search | How it Works and Sklearn Implementation

This series is going to focus on one important aspect of ML, hyperparameter tuning. In this video we are going to talk about grid search, including what it is and how to use the scikit-learn GridSearchCV module. In the module, we are going to dive into some important fields that you might want to customize, instead of just using the default settings, and these are going to include the scoring method, CV splitting method, and how to choose the best set of hyperparameters after fitting a GridSearchCV.

Timestamps:
Why the name Grid Search, some intuition, and how it works: 0:48
3 core questions to ask before starting a grid search: 1:47
Explaining key inputs to the sklearn GridSearchCV module: 2:14
Explaining cross validation: 4:01
Simple implementation: 4:47
How to read the cv_results_: 6:10
Customize cross-validation: 7:26

Documentation links:
Using different metrics: https://scikit-learn.org/stable/modul...
GridSearchCV: https://scikit-learn.org/stable/modul...

Credits to Xoel López Barata for the interactive visualization plot, the original article at Towards Data Science:
https://towardsdatascience.com/using-...