A fantastic video that explains catboost so well! I wish I had seen it before the last tech interview...

Thanks!! Would be very appreciated if you could do a video in which you implement CatBoost with Cross Validation in python!!

Hey, nice explanation on how catboost does categorical encoding during training which makes it robust to overfitting and also faster but there is another very important feature 'ordered boosting' of catboost which i think you could cover. I was reading the paper but couldn't grasp it properly. Anyways thanks for the video :)

yes plz make video on python implementation of lightgbm and cateboost .. with time series implementation and hyper parameter implementation

Hello Aman, thanks for the sharing the details I have quick question on data encoding for Nominal data, which one will select among list of the encoding techniques ? One-hot/dummy encoding. Label / Ordinal encoding. Target encoding. Frequency / count encoding. Binary encoding. Feature Hashing.

This is truly a fantastic explanation!

Hi Aman, thanks for sharing the video. nice explanation. Helps a lot. One question on categorical var encoding, i understand the training data part (you made it easy!) but curious to know how it is done in test set. Is it mean/median of the numeric encoded values for each category of a category var in the training set? Thanks

Thanks for the New concept. Can we use this Catboost for Text Classification and Time Series. And please upload the Python code implementation for the Catboost.

What do you mean by Data leakage and how does it affect the model performance if we take the average of all values from the dataset as you mentioned at 10:50

can we also expect practical implementation for the same..

Nice explanation. Thanks Aman.

Please create a video of cat boost mathematics

Hi Aman, if encoding is different for every row how the category will be encoded at the time of prediction.

Can u please tell me what is purpose of using catboostregression ?

What writing you are using ?

Well explained

can catboost and lightgbm algo used for time series forecasting plz tell me ?

If my data consists of both categorical and numerical feature can i implement catboost directly . coz in the error it says that it has to be str or an int only

if threre is no cator. variable then we dont need to do pooling ??

Broi..is this relevant outside russia??