Principal Component Analysis (PCA) | Dimensionality Reduction Techniques (2/5)

[@English-bh1ng]

Well-organized video and description, abundant references. I love this series. Cheer up!

[@yashdevarshi2583]

Amazing explanation! Wish my uni professors were like you

[@anas.aldadi]

what a nice a cool channel to discover, i stumpled upon your channel searching for mathematical explaination for diffusion theory and model!

[@DeepFindr]

Thanks! Appreciated :)

[@MegaBoss1980]

In your future series, will you also cover PCA for categorical variables? Also, can we apply PCA on embeddings of categorical variables?

[@DeepFindr]

Hi :) for this series that's the only thing about PCA. Next videos will be about other techniques. It's mainly intended to get a good overview for each method. PCA is designed for continuous variables - all of the projections don't make too much sense for categorical data. That's mainly because distances are not properly defined. Of course it's possible to apply it anyways for example on one hot encoded variables, but it might not be the best choice. You might want to look into Multiple Correspondence Analysis (MCA), which is designed for categorical variables.

[@TheZapalsky]

great content!

[@shaz-z506]

Nice one 😃, could you please extend this and explain kernel PCA in the similar manner, I don't think so there are many videos kernel PCA

[@DeepFindr]

Will put it on the list but can't promise :D

[@yeshiwangmo5920]

Do you have ppt on this

[@FabioDBB]

Amazing explanation dude Rome is way bigger than NYC btw

[@DeepFindr]

Thanks! Yeah area-wise it is but not population-wise, right?

[@YouKnowWhoIAm118]

Hi, your explainable AI playlist could be updated ;) no offense bro, just as a suggestion

[@DeepFindr]

Hehe with which method? :)

[@lorenzoneri-co5hj]

(rome is bigger than nyc)

[@DeepFindr]

When it comes to area probably yes :P but not citizens wise