catboost explained | catboost algorithm explained | catboost vs lightgbm vs xgboost

[@miaojunlee]

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

[@omprakashverma6549]

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 :)

[@UnfoldDataScience]

Welcome OM 🙂

[@mohitmittal2627]

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

[@UnfoldDataScience]

Video is created Mohit. Please look on the channel for "xgboost vs catboost vs lightgbm"

[@tnmyk_]

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

[@UnfoldDataScience]

Watch thois Tanmay kzbin.info/www/bejne/r6aygKJme7aeorM

[@exxyyf]

This is truly a fantastic explanation!

[@simonebenzi4189]

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

[@bijaynayak6473]

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.

[@ajaykushwaha-je6mw]

can we also expect practical implementation for the same..

[@UnfoldDataScience]

It's there, search for catboost vs lightgbm vs xgboost

[@ankursharma1836]

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

[@anindyasil6099]

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

[@dhamodharanr3461]

What writing you are using ?

[@tanmaychakraborty5472]

Please create a video of cat boost mathematics

[@karthebans248]

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.

[@UnfoldDataScience]

Welcome. Yes we can use. Sure

[@radhakrishnanananthan1585]

Nice explanation. Thanks Aman.

[@UnfoldDataScience]

Welcome Radha.

[@pareshdashore2772]

Well explained

[@UnfoldDataScience]

Thank you

[@husnakhan7683]

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

[@UnfoldDataScience]

In the pipeline mention which ones are categorical columns. Check my implementation video, I hv done it. Video title "catboost vs xhboost vs lightgbm"

[@tanyagupta6319]

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

[@mohitmittal2627]

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

[@UnfoldDataScience]

Yes possible

[@mysteriousocean6198]

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

[@UnfoldDataScience]

I dont think so

[@bibinkunjumon]

Broi..is this relevant outside russia??