- notebook from the session: www.kaggle.com/robikscube/handling-with-missing-data-youtube-stream - Rob's Kaggle: www.kaggle.com/robikscube - Rob's twitch: www.twitch.tv/medallionstallion_ - Rob's twitter: twitter.com/Rob_Mulla - Rob's youtube: kzfaq.info/love/xladMszXan-jfgzyeIMyvw - pog champ kaggle competition series: www.kaggle.com/c/kaggle-pog-series-s01e01

This channel is a blessing to Data science people

Very good video, I didn't know this Kaggle grandmaster, I loved the way he explained everything. I hope you continue collaborating with each other, your knowledge is super valuable for people like me who are just starting out. Greetings!

So much good info in this video. Rob explained it really well. Enjoyed his initial data viz of how the missing values get filled with mean, zeroes, ffill etc; it is such a handy way to show how changes manipulates the data. Thanks for this video.

Thanks for the video. Very informative! I'd like to have more streams with Rob. He is great at delivering knowledge and definitely has a lot to share

Thank you so much for this amazing walk through. Learnt a lot. Super helpful. 👌

This is all I have been looking for, thanks a lot.

Thank you Abhishek and Rob. It was very informative session.

Rob Mulla, a humble man of God.

Thank you both, we learned new stuff today!!

Thank You for such a wonderful session

Learnt a lot, thanks.

Thanks for the content!

ty so much!

First time I see the query function! Why would you prefer it instead of using the typical way? (E.g. ncounts[ncounts["train_missing > 0"]] )

Can somebody explain what Rob is talking about in 1:00:13 ? Why are we taking mean of test imputer predictions ?

Nice!!

I am getting value error while using LGBM imputer. So does anyone have any idea that what kind of features it accepts?