Basics of PCA (Principal Component Analysis) : Data Science Concepts

No video

Basics of PCA (Principal Component Analysis) : Data Science Concepts

Рет қаралды 99,408

5 жыл бұрын

Gentle Intro to Principal Component Analysis (PCA)
---
Like, Subscribe, and Hit that Bell to get all the latest videos from ritvikmath ~
---
Check out my Medium:
/ ritvikmathematics

Пікірлер: 70

@jorostuff 3 жыл бұрын

Finally, someone who introduces the problem first, the need, and then shows the solution. Most videos give you the solution and don't give you the problem at all, so you're left wondering why the thing exists in the first place. Great video!

@AJ-ks8iq 3 жыл бұрын

The best comment I have read in forever.

@azamatbagatov4933 2 жыл бұрын

Better than Josh Starmer

@sanketannadate4407 2 ай бұрын

Really Great comment, coulnt agree more

@r0cketRacoon 16 күн бұрын

@@azamatbagatov4933 ? Josh Starmer also does that

@isman32 3 жыл бұрын

This "high level" explanation was just what I needed to get rid out of doubts about how to implement PCA. Thanks so much for your effort. You got 1 new follower.

@ritvikmath 3 жыл бұрын

Glad it helped!

@Ali_Alhawaj 4 жыл бұрын

Finally, a clear video on PCA for a guy coming from biomedicine. Thank you!

@AJ-ks8iq 3 жыл бұрын

I don't know if you realize how good you are!

@waynethompson8314 9 ай бұрын

I started teaching business analytics and find your teaching method as very consumable. Assuming length and weight of kitties is correlated I assume you could have just one PC.

@robertc6343 3 жыл бұрын

You have a rear gift of explaining things in a very clear way my friend. I’m curious to see the next video!

@ritvikmath 3 жыл бұрын

Thank you very much!

@niamhtromans9086 3 жыл бұрын

Finishing a degree during COVID-19 and this man has explained better in 6 minutes what a lecturer I'm paying £9,000 spent 3 different 1 hour teams meetings trying and failing to explain

@sammyj29 3 жыл бұрын

One of the best channels which go in depth while still keeping it clear. Thank you for creating such content!

@AJ-ks8iq 3 жыл бұрын

and not bore you with the math, if you don't need it.

@Information-Overlord 2 ай бұрын

All the great reviews below are understatements ;-)

@zakzeaxer4868 3 жыл бұрын

looked a bunch of videos, but only this explained me well the interest of PCA !

@frenchmike 6 ай бұрын

thanks for explaining it in plain english before going into some crazy math. I appreciate this

@sulaiman.micheal Жыл бұрын

You just saved me a lot of time and stress with this simple straightforward film. Thanks a lot. Please make another one like this on the fisher-face and LDA algorithm. 😇😇😇

@ritvikmath Жыл бұрын

Glad it helped!

@noahrubin375 3 жыл бұрын

This was outstanding. Exactly what I needed.

@priyankagupta119 2 жыл бұрын

When we study, our aim is to get clarity where we will use this concept, you explained it very well. Thank you so much.

@armonnaddaf5326 2 жыл бұрын

Hey Ritvik, I might be your biggest fan! Whether I'm cramming for an exam, trying to understand a paper, or just killing time on a curious Saturday, I always look for your videos first! That said, it would be very helpful if you could include a link to part 2 in your description or at the end of the video. Thank you for all you do :)

@viviann3540 4 жыл бұрын

Really like your explanation of PCA! I am a food microbiology Ph.D. student. Thank you for this great video! Hope more people can see this！Cheers！

@ritvikmath 4 жыл бұрын

Glad it was helpful!

@francescodipietro4340 6 ай бұрын

Thanks man, i really need this kind of video to understand the topic

@fionadu-chan9531 6 ай бұрын

You are a genius!

@guilhermehx7159 Жыл бұрын

Very insightful

@ElectricTeaCup 3 жыл бұрын

"We are cat researchers" *Instantly likes*

@ritvikmath 3 жыл бұрын

@user-xj4gg9jm3q 2 жыл бұрын

fantastic video, the examples and using real numbers really clarifies things tysm !!

@salamkanada 3 жыл бұрын

Great video! Clear and well explained. Really appreciate it. Thank you

@ritvikmath 3 жыл бұрын

Glad it was helpful!

@hameddadgour Жыл бұрын

Great video! Thank you for sharing.

@sethquigley3677 2 жыл бұрын

Awesome work man! Thank you and God bless

@souravdey1227 3 жыл бұрын

Absolutely genius. I was looking for an intuitive understanding and here you se brother. Can you request you to do a few videos on locality preserving projection?

@user-tb5kz3mx9g 7 ай бұрын

Very well explained. Thanks

@ritvikmath 7 ай бұрын

Glad it was helpful!

@wasimasadia4887 Жыл бұрын

Nicely explained. Thank you.

@sArAsara-rr2qm Жыл бұрын

this video helpt me a lot thank u

@_Sam_-zh7sw 9 күн бұрын

Bro. your videos are amazing and we really appreciate your work. But we are dependent on your channel so can you structure your content into bins of playlists? Or may be just make one playlist but arrange all videos in proper sequence.

@garcon2008 Жыл бұрын

This was excellent.

@mehdicharife2335 2 жыл бұрын

Thank you.

@manulwild6967 4 жыл бұрын

Great explanation, thank you!

@firstkaransingh 2 жыл бұрын

Can you do a video comparing V.I.F and PCA and LASSO/RIDGE regularization. They seem to deal with very similar objective of multicollinearity. Excellent explanation as always.

@Paprika526 3 жыл бұрын

This probably hurts me more than it should, but he means purr volume instead of purr frequency. Purr frequency is either the amount of times a cat purrs, or the tonal frequency on which the cat purrs. Yes I'm fun at parties.

@ritvikmath 3 жыл бұрын

love it!

@teetanrobotics5363 2 жыл бұрын

Amazing content. Could you please order the playlist on your channel for beginners ?

@ankitranjan30 2 жыл бұрын

Thank you for the video, Ritvik! Not related to PCA but I was wondering if you could make a video on latent dirichlet allocation (lda) for topic modeling? It would be really helpful.

@Reivivus 2 жыл бұрын

Liked and subscribed

@pickinnick100 4 жыл бұрын

Frequency of purr =/= Amplitude of purr

@ajiths1689 4 жыл бұрын

great explanation

@HA-vh3ti 3 жыл бұрын

How would you compare PCA vs Variable Selection/Regularization (Lasso/Ridge/Elastic Net), Is there any video comparing both?

@velevki 2 жыл бұрын

We are saying that we are reducing the dimensions but each PCA is a linear combination of all the attributes. So how is this reducing dimension? This is the part I am not grasping.

@smudgepost Жыл бұрын

Very interesting thank you

@ritvikmath Жыл бұрын

Glad you enjoyed it

@smudgepost Жыл бұрын

@@ritvikmath I'm really enjoying your videos because it's giving me a working understanding of the mechanics under the hood. I've been toying with data science for years and struggled quite a bit, but with videos like this and the aid of AI bots to help (particularly with tracing errors), I'm learning so much faster today. As you get into great detail across the whole spectrum of machine learning I wondered if you had a video or might consider making one to show the structured approach to a machine learning project. Looking at the type of data the desired outcome and then piece-mealing all of the tools and approaches to the strategy. Each of your videos is very holistic, I would love to see one that was holistic for a whole machine learning project, without needing to re-explain the detailed concepts held in your other videos.

@tirthupadhyaya2519 3 жыл бұрын

You are awesome dude

@mirzaaghaalikhan183 3 жыл бұрын

I was having a very bad day and this showed up! *Thanks KZfaq*

@davidcrawley9479 4 жыл бұрын

Terrific video by the way. Do you mean you loose a little bit of data or do you mean you loose a little bit of information? I am guessing that what you are saying is that you loose data without loosing information.

@beshosamir8978 2 жыл бұрын

how can i know, how much every PCA component capturing data ?

@sahil0094 2 жыл бұрын

how is dimensionality reduction & feature extraction different?

@iaaan1245 4 жыл бұрын

does PCA solve the issue of multicollinearity?

@theonionpirate1076 2 жыл бұрын

"A little messy, but it says 'feature extraction'." Agree to disagree.

@DeAdAnGeLTV 4 жыл бұрын

You say your cat dataset has 3 Dimensions where as one could be left out because it has no great significance. I know it is a toy example but why wouldn't you just leave it out without calculating it with PCA?

@ritvikmath 4 жыл бұрын

Great question. In this case when it is easy to visualize in 3d, we can just leave out that dimension. But, when the dimensions get higher and it get's harder to visualize you might consider PCA. The bigger reason is that sometimes the points all lie on a line or plane but that line or plane does not line up perfectly with the axes (one variable is a linear combination of others) so PCA is needed.