Hyperparameter Tuning for Machine Learning Models - ML 079
Adventures in Machine Learning
00:00
Using Grid Search for Decision Trees and Random Forests
If we're given a range between zero and one, what are we always going to select? We're either going to generate a range by 0.1 to say, Hey, here's my 10 things. And then that's sort of one problem is you're selecting something that is assuming a linear relationship between its scale and the response. So I'm always a big fan of random search. It finds better parameters than manual grid search unless you're doing like 10,000 searches through grid search.
Transcript
Play full episode
