Have you been using grid search to tune your models (or pipelines)? In this video, you'll find out why randomized search is sometimes a better fit! Feel free to let me know if you have any questions 💬
Great video. I use Randomsearch to get in the 'ballpark' hyperparameter-wise, basicially some starting values of which I can then use Gridsearch to fine-tune around the Randomsearch-recommendedations and zero-in on the optimal hyperparameters. Probably overkill but it seems to work for me. Keep up the great content.
Hello, I really like the high quality and clearly explained content! I'm trying to do feature selection where I have mostly categorical, non ordered inputs, some numerical inputs, and a numerical output. I've been researching everywhere for the past few days, but I can't seem to find a resource/guide that talks about my specific use case, do you know any places I can go to learn how to do this?
Thanks for your question and your kind comment! I have a chapter on feature selection in my upcoming ML course, but unfortunately I haven't seen this topic covered in a clear and thorough way in other courses, and mine won't be available for a few more months. The best resource I can think of at the moment is this lesson: inria.github.io/scikit-learn-mooc/feature_selection/feature_selection_module_intro.html
Great video. Hi, I'am trying to use the RandomizedSearchCV to find the best parameters for a clustering method. I'am using the make_score with the attribute great_is_better and passing the validity_index, but it gives a message that the all estimators failed. Do you have any video teaching how to use the RandomizedSearchCV for unsurpervised models? Tkx
Hi i am new to this field, i was trying to apply Grid Search and RandomizedSearch to my model but I am not able to understand the parameters that people talk about an where can i get a list of what can be tuned in a specific model that I am using
Have you been using grid search to tune your models (or pipelines)? In this video, you'll find out why randomized search is sometimes a better fit! Feel free to let me know if you have any questions 💬
Great video. I use Randomsearch to get in the 'ballpark' hyperparameter-wise, basicially some starting values of which I can then use Gridsearch to fine-tune around the Randomsearch-recommendedations and zero-in on the optimal hyperparameters. Probably overkill but it seems to work for me. Keep up the great content.
Excellent idea, thanks for sharing!
Great video! this save my time a lot😊
Great to hear!
Hello, I really like the high quality and clearly explained content! I'm trying to do feature selection where I have mostly categorical, non ordered inputs, some numerical inputs, and a numerical output. I've been researching everywhere for the past few days, but I can't seem to find a resource/guide that talks about my specific use case, do you know any places I can go to learn how to do this?
Thanks for your question and your kind comment! I have a chapter on feature selection in my upcoming ML course, but unfortunately I haven't seen this topic covered in a clear and thorough way in other courses, and mine won't be available for a few more months. The best resource I can think of at the moment is this lesson: inria.github.io/scikit-learn-mooc/feature_selection/feature_selection_module_intro.html
Great Video 👍🏻. It would be really helpful if you could make a video about HyperOpt or Optuna. Thanks.
Thanks for your kind words and your suggestion!
Great video.
Hi, I'am trying to use the RandomizedSearchCV to find the best parameters for a clustering method. I'am using the make_score with the attribute great_is_better and passing the validity_index, but it gives a message that the all estimators failed. Do you have any video teaching how to use the RandomizedSearchCV for unsurpervised models?
Tkx
thank you so much. i thought i was going crazy, everytime i do gridsearch, it takes forever. i will try this out
Great!
Hi i am new to this field, i was trying to apply Grid Search and RandomizedSearch to my model but I am not able to understand the parameters that people talk about an where can i get a list of what can be tuned in a specific model that I am using
Great question! The short answer is to read the user guide and API documentation for scikit-learn for the particular model you are using.
Really it helps a lot
Great to hear!