I learned about grid search for a course last semester, it was for a final project on sklearn that I waited to start 3 days before it was due, so I rushed to learn a lot of concepts that I promised myself I’d go back and understand more thoroughly (and probably will need to for upcoming courses/career). So this was an awesome refresher on this topic! Love your videos, keep up the great work!
DUDE!!! You are the Guru among common minds. This was the best explanation ever. Simple succinct and easily understandable for a newbie like me. I also like you give extra homework, and nuggets of knowledge related to the topic that I can look into afterwards. Now time to study the rest of your videos 👍⭐⭐⭐⭐⭐
This is my first comment at youtube. I came here because of your video quality and realize your explanation also fanstastic. I think you are using manim. For me it's a great chanel.
I knew about grid search, but in the end the technique you shared was new to me and it will be really helpful in the future when I'm going to use grid search, thank you for sharing. Great vid as always 😊
I have just stumbled upon you channel. you videos are well communicated. Well done. Another good video could be - how do you settle on a model that is generalised to test data.
Thank you. Please can you make a video explaining how we can make prediction from different regression models: regression tree, random forest, artificial neural network, SVM, Bagged CART, Generalized boosting, Extreme Gradient boosting
Thanks. Nice video. But why are you doing GS.fit (X_train, Y_train) when you already have cross validation with cv = 5. Shouldn't you just do GS.fit (X, Y) ?
you should train on x_train and y_train because this allows you to further test the model once the optimal hyperparameters have been found with an unseen test set which the model has not seen stopping data-leakage
thanks. Can you make a video about deep neural network regression predictions. Like multiple inputs and one output prediction by using this gridsearch. Thanks a lot.
Thanks, one queation, after tuining the model and finding the best hyper parameter, is it necessary to run the model with found best parameter for moel training right and prediction? i mean after using GridSearchCV, model is already configured by best parameter? can you please elaborate?
thank you for the video it was really helpful. I had a question, could you please help me? when I want to download data from kaggle, I receive 403 forbidden Error!
Thank you for this beautiful work. I have a question: what happens if we run the same code several times? Will the best_estimator always be the same? Thank you for your answer.
Bro, even without hyper parameter tuning, I am getting more than 0.99 r2_score. But I am getting 0.96 r2_score with tuning. So how exactly this tuning is helpful?
I learned about grid search for a course last semester, it was for a final project on sklearn that I waited to start 3 days before it was due, so I rushed to learn a lot of concepts that I promised myself I’d go back and understand more thoroughly (and probably will need to for upcoming courses/career). So this was an awesome refresher on this topic! Love your videos, keep up the great work!
Thanks mate! It's always useful to revisit an old topic.
DUDE!!! You are the Guru among common minds. This was the best explanation ever. Simple succinct and easily understandable for a newbie like me. I also like you give extra homework, and nuggets of knowledge related to the topic that I can look into afterwards. Now time to study the rest of your videos 👍⭐⭐⭐⭐⭐
This is my first comment at youtube. I came here because of your video quality and realize your explanation also fanstastic. I think you are using manim. For me it's a great chanel.
Welcome to my channel! I'm really glad you that liked my content :D :D
Yes, I use manim.
Hope you are doing well sir . Kindly continue your video series as it greatly helps and is just amazing!! 🙂
Very nice and one of the best video on Hyperparameter Tuning
Thank you!! This is the best video on this subject!
thanks for your work and dedication . your vid is very useful for my final project in DS bootcamp.
I knew about grid search, but in the end the technique you shared was new to me and it will be really helpful in the future when I'm going to use grid search, thank you for sharing. Great vid as always 😊
Yeah that's really helpful.
Great video, thanks again!
Glad you enjoyed it!
Recently found your channel very grateful
Nice, clean animations 👍
amazing video my friend!
I have just stumbled upon you channel. you videos are well communicated. Well done. Another good video could be - how do you settle on a model that is generalised to test data.
Thanks! Gonna try this out!
keep up the good work..😀
Awesome Explaination! appreciate your work and subscribed;)
great video helped me alot!
thanks for sharing especially for the last part that how to choose an efficient model with low computer resource wasting haha
Glad it was helpful!
Great content. Thanks
Thank you. Please can you make a video explaining how we can make prediction from different regression models: regression tree, random forest, artificial neural network, SVM, Bagged CART, Generalized boosting, Extreme Gradient boosting
Great video. thank you
thank you so much for the great video
This is awesome. Thanks!
Awesome video. tnx man!!
Thanks mate!
Thanks. Nice video. But why are you doing GS.fit (X_train, Y_train) when you already have cross validation with cv = 5. Shouldn't you just do GS.fit (X, Y) ?
you should train on x_train and y_train because this allows you to further test the model once the optimal hyperparameters have been found with an unseen test set which the model has not seen stopping data-leakage
Best one! Thanks!!
Amazing! How can I implement this on fastercnns?
Great vid! Thanks!
Thanks mate! :D
Hello, congratulations for video.
I have a question. You have a example that discovers the best hyper-parameters using Swarm Intelligence?
thanks. Can you make a video about deep neural network regression predictions. Like multiple inputs and one output prediction by using this gridsearch. Thanks a lot.
Why do we need to split it to be X_train and y_train when internally in gridsearchCV we will conduct k-fold? please help me answering this question.
Thanks, one queation, after tuining the model and finding the best hyper parameter, is it necessary to run the model with found best parameter for moel training right and prediction? i mean after using GridSearchCV, model is already configured by best parameter? can you please elaborate?
Could you do a video about boosting models please? Thanks
thanks , my friend
Thank you bro
ty
Thank you!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Thanks for your video it is going to help me ! :D
Invalid parameter "max_depth " contains whitespace.
I got error kind of this ..
thank you for the video it was really helpful. I had a question, could you please help me? when I want to download data from kaggle, I receive 403 forbidden Error!
Thank you for this beautiful work.
I have a question: what happens if we run the same code several times? Will the best_estimator always be the same?
Thank you for your answer.
if you keep random_state the same number, I guess there will be the same best_estimator
great
Bro, even without hyper parameter tuning, I am getting more than 0.99 r2_score. But I am getting 0.96 r2_score with tuning. So how exactly this tuning is helpful?
it means default values are best fit for your dataset
What about the the testing data? It seems u have not used them.
How does GridSearch work with pickeld data?
From now on, I identify as a person from the future😊
which software do you use for your animation?
I used manim for intro animation
@@NormalizedNerd and for the rest of the video?
there'll alway be a random indian out there who will help you get your assignment done on time.
lol...at least you are not outsourcing your assignments!
Where are you? We miss you :(
1:53 -> 2D arrays not lists 😀
Hey buddy how old are your 😁
Random state why = 2021
Nice