PCA Indepth Geometric And Mathematical InDepth Intuition ML Algorithms

แชร์
ฝัง
  • เผยแพร่เมื่อ 7 ก.พ. 2025
  • github Materials: github.com/kri...
    Principal component analysis (PCA) is a popular technique for analyzing large datasets containing a high number of dimensions/features per observation, increasing the interpretability of data while preserving the maximum amount of information, and enabling the visualization of multidimensional data. Formally, PCA is a statistical technique for reducing the dimensionality of a dataset. This is accomplished by linearly transforming the data into a new coordinate system where (most of) the variation in the data can be described with fewer dimensions than the initial data. Many studies use the first two principal components in order to plot the data in two dimensions and to visually identify clusters of closely related data points. Principal component analysis has applications in many fields such as population genetics, microbiome studies, and atmospheric science.

ความคิดเห็น • 93

  • @exploreEverything4519
    @exploreEverything4519 ปีที่แล้ว +25

    First I understood pca concept 3 years back from nptel lecture. It was full of mathematics and It went far above my head because the theory part was missing. Believe me with your explanations I can understand his lecture too. No one could explain the way you have explained. It was outstanding.

  • @aditinautiyal4299
    @aditinautiyal4299 ปีที่แล้ว +7

    Thank you so much for not only sharing your knowledge but also putting so much effort to cover each and every point of the particular topic.

  • @salminsultana9654
    @salminsultana9654 25 วันที่ผ่านมา +1

    So I am your new fan sir.
    Today I was searching for videos about the auto eda.
    And I found yours. After that I started to watch your videos about pca clustering etc. And you are just amazing.. now I can understand properly. Thanks a lot sir.

  • @IshanGarg-y1u
    @IshanGarg-y1u ปีที่แล้ว +14

    This is a good video, I recommend first you watch PCS step by step guide from stat quest to get a high level view with animations, then you watch this video to get more details and understanding alongside some code. Then in case you want to know the mathematics behind it refer to some articles online where the explain why we calculate the covariance matrix, then build the objective function using lagrange multiplier and then derive why eigen values of covariance matrix are the desired results

  • @aj_actuarial_ca
    @aj_actuarial_ca ปีที่แล้ว +1

    PCA is so very well explained in your video sir. You're really the best teacher ever !!!

  • @pritamrajbhar9504
    @pritamrajbhar9504 9 หลายเดือนก่อน +4

    thanks a lot, Krish this is the simplest and most detailed video about PCA.

  • @akashpaul9892
    @akashpaul9892 ปีที่แล้ว +2

    You really are a good teacher brother... Teaching with relatable examples help to understand each topic so perfectly and easily.. Thank you so much brother.. Keep teaching us...
    Love from Bangladesh

  • @man9mj
    @man9mj 11 หลายเดือนก่อน +1

    thank you for this elegant effort in explaining PCA

  • @Harsh_Yadav_IITKGP
    @Harsh_Yadav_IITKGP ปีที่แล้ว +1

    Krish your efforts are remarkable in this ml series.....

  • @syco-brain8543
    @syco-brain8543 4 หลายเดือนก่อน +1

    best video about pca on internet so far

  • @yogendrapratap1982
    @yogendrapratap1982 ปีที่แล้ว +16

    Everything had been really resourceful in lecture series but this lecture was overly extended, 30 min topic has been extended to 1 hours 30 mins repeating same stuff again and again

  • @taslima5007
    @taslima5007 10 หลายเดือนก่อน

    You are my favourite youtuber and teacher.

  • @adnanshujah6230
    @adnanshujah6230 9 หลายเดือนก่อน

    best of the best lecture .covers all the required concepts about subject . most of videos available only shows how to perform PCA but not whay it is required and concept behind it .but sir Krish thankyou so much for such a detailed lecture and clearing the concepts . highly recommended lecture and his channel
    🥰🥰🥰🥰🥰🥰

    • @adnanshujah6230
      @adnanshujah6230 9 หลายเดือนก่อน

      i simply say this one video is enough to get the clear concept ;once again thankyou soooooo .... much sir Krish

  • @SanthoshKumar-dk8vs
    @SanthoshKumar-dk8vs 2 ปีที่แล้ว +3

    Thanks for sharing Krish really helpfull, last two days am refreshing this topic only🤗

  • @ashwintiwari9642
    @ashwintiwari9642 2 ปีที่แล้ว

    No where I can find this explanation it's too good no confusion no complex demonstration use cases a cleanest and simplest way to understand PCA in depth thanks alot Krish it takes lot of takes and research to explain single topics in data science and in this way it's all appreciated work

  • @dipamsarkar6626
    @dipamsarkar6626 ปีที่แล้ว +1

    This guy should be named as "God father of Data Science India" an absolute legend

  • @BhagatRaviNarayan
    @BhagatRaviNarayan 3 วันที่ผ่านมา

    He used the comparison operator in reverse order Accuracy1 < Accuracy2 < Accuracy3 > Accuracy4 > Accuracy5, Due to curse of dimensionality. 5:35

  • @vinothkumar7531
    @vinothkumar7531 ปีที่แล้ว

    You are a great teacher I ever seen in my entire life.The way you are teaching even makes the lazy or slow learner to a strong learner using Krish Naik g(ji) Boosting algorithm.Just Kidding 😃😃.Hatsoff to your effort to help the people.

  • @amitx26
    @amitx26 10 หลายเดือนก่อน

    Sir, I thing have felt strongly is that you expain and deliver a little better in recorded videos. Thanks for providing such great content for us for free!

    • @RakshithML-vo1tr
      @RakshithML-vo1tr 10 หลายเดือนก่อน

      Hi bro I am starting data science how can I start? By seeing Krish sir roadmap and like u said should I prefer recorded videos

  • @paneercheeseparatha
    @paneercheeseparatha ปีที่แล้ว +4

    Wonderful try to explain PCA without much mathematics. Though it would be great if you also do a video on implementing PCA from scratch in python. Loved your playlist! kudos to you!

  • @samareshms4591
    @samareshms4591 10 หลายเดือนก่อน +3

    This guy is single handedly carrying the AI ML community in the India 🙇‍♂🙇‍♂

  • @viratkumar9161
    @viratkumar9161 ปีที่แล้ว +2

    Its quite vage to say if pearson correlation value is zero there is no relationship between x and y. Example consider Y= mod(X) line the person correlation is 0, but still there is relationship easily visible after plotting

    • @SiddharthSwamynathan
      @SiddharthSwamynathan ปีที่แล้ว

      Correct. Pearson correlation has the capacity only to capture the linear relationship. Coefficient 0, would be no linear relationship exists. But there exists a possibility of a non linear relationship within the covariates and target.

  • @ramakrishnayellela7455
    @ramakrishnayellela7455 9 หลายเดือนก่อน

    Such a good explanation krish

  • @kvafsu225
    @kvafsu225 ปีที่แล้ว

    Excellent presentation.

  • @irisshutterwork1411
    @irisshutterwork1411 ปีที่แล้ว +1

    Well explained. Thank you

  • @meghananddussa62
    @meghananddussa62 16 วันที่ผ่านมา

    Covariance(direction) , correlation coefficient (strength and direction) both will tell about linear relationship only, they do not explain about non-linear relationship

  • @pankajray5939
    @pankajray5939 2 ปีที่แล้ว +1

    PCA is one of the important topics of ML

  • @baravind6548
    @baravind6548 8 หลายเดือนก่อน +2

    In extracting from 2D to 1D, if PC1 has the higer varience and PC2 has 2nd higher varience. Is it nessesary that PC1 should be perpendicular to PC2?

  • @thop9747
    @thop9747 2 ปีที่แล้ว

    was really helpful. Keep up the work sir.

  • @the-ghost-in-the-machine1108
    @the-ghost-in-the-machine1108 ปีที่แล้ว

    Thanks sir, god bless you!

  • @harshitsamdhani1708
    @harshitsamdhani1708 ปีที่แล้ว

    Thank You for the video

  • @unicornsolutiongh2022
    @unicornsolutiongh2022 ปีที่แล้ว

    powerfull lecture. keep it up sir

  • @manikandanm3277
    @manikandanm3277 2 ปีที่แล้ว +3

    In theory part, to find the eigen values, you multiply the covariance matrix with a vector. How's that particular vector V is chosen and used to multiply with the covariance matrix? I'm confused with this only, otherwise a great lecture, thanks krish👍

    • @priyam39
      @priyam39 ปีที่แล้ว

      That v is the eigen vector itself we are looking for.Sir just explained

  • @sumankumar01
    @sumankumar01 ปีที่แล้ว +1

    Campus x and you both refer same books or what since the example is same ?

  • @bhagyashriakolkar7763
    @bhagyashriakolkar7763 ปีที่แล้ว

    Thank you sir....nice explanation

  • @mr.pianist
    @mr.pianist 6 หลายเดือนก่อน

    very good lec beginner friendly

  • @AjayPatel-pc1yf
    @AjayPatel-pc1yf ปีที่แล้ว

    Gjb sir mja aa gaya❤

  • @yachitmahajan3579
    @yachitmahajan3579 10 หลายเดือนก่อน

    best explanation

  • @Nikhillllllllllllll
    @Nikhillllllllllllll ปีที่แล้ว +1

    how to get names of those 2 features we got after feature extraction

  • @CodeWonders_
    @CodeWonders_ 2 ปีที่แล้ว +1

    Can you tell me who will teach in data science course you or sudhanshu sir ?

  • @javeedtech
    @javeedtech 2 ปีที่แล้ว

    Thanks for video, from fsds batch 2

  • @shivachauhan2837
    @shivachauhan2837 2 ปีที่แล้ว +2

    To improve my resume what should I try kaggle Or open source

  • @lagangupta3193
    @lagangupta3193 6 หลายเดือนก่อน

    How will we decide the number of features that we have to mention in n_components?

  • @RahulA-b9o
    @RahulA-b9o ปีที่แล้ว

    How do i know that the model is over feeded.. any method to find out that the model trained is under curse of Dimensionality???????

  • @IzuchukwuOkafor-v6e
    @IzuchukwuOkafor-v6e 11 หลายเดือนก่อน

    Very lucid explanation of PCA.

  • @PAVVamshhiKrishna
    @PAVVamshhiKrishna 5 หลายเดือนก่อน

    Fantastic

  • @jitendrasahay3847
    @jitendrasahay3847 4 หลายเดือนก่อน

    If we have 3 features then we are getting 3 eigen vectors and later we combine 2 out of them to create 1 eigen vector. Combining here basically mean projection. Earlier when we projected we got n eigen vectors out of n feature then again we will get 2 eigen vectors. Where the dimensionality reduction is happening???
    What I m missing here really???
    Can anyone help ???

  • @ITSimplifiedinHINDI
    @ITSimplifiedinHINDI 7 หลายเดือนก่อน +1

    Greater than ko Less than aur Less Than ko Greater Than, kyoun likh rahe ho Guruji.

  • @Bitter_Truth-zc4eq
    @Bitter_Truth-zc4eq ปีที่แล้ว

    Which software are you using for writing?

    • @KRSandeep
      @KRSandeep 10 หลายเดือนก่อน

      Scrble Ink which is available for windows laptop only

  • @samthomas3881
    @samthomas3881 11 หลายเดือนก่อน

    Thanks Sir!

  • @user-rx5kq6oo9y
    @user-rx5kq6oo9y 2 ปีที่แล้ว +4

    Bro can you make cheat sheet of data science like multiple dsa
    sheets on youtube?

  • @eurekad7340
    @eurekad7340 4 หลายเดือนก่อน +1

    If possible could you please make video on truncated svd as well. I searched but I couldn't find any video on svd from you

    • @faizannaseem3384
      @faizannaseem3384 4 หลายเดือนก่อน

      See Go Classes Free Leactures for SVD

  • @BharatDhungana-n4s
    @BharatDhungana-n4s ปีที่แล้ว

    implementation is best

  • @MamunKhan-px2vb
    @MamunKhan-px2vb 2 ปีที่แล้ว

    Just Great

    • @MrKhan-xu1vf
      @MrKhan-xu1vf 2 ปีที่แล้ว

      Kinda amazing teaching skills

  • @kunalpandya8468
    @kunalpandya8468 ปีที่แล้ว

    After we get 2 features from pca, what is the name of those two features?

  • @chayanikaboruha6657
    @chayanikaboruha6657 11 หลายเดือนก่อน

    Krish please make a video regarding how we can use auto encoder for text data

  • @baravind6548
    @baravind6548 8 หลายเดือนก่อน

    How to get the vector v? that is to be multiplied by A

  • @BMVLM-
    @BMVLM- 2 หลายเดือนก่อน

    Bhai content mast hai lekin advertisment bhot sare hai bot disturbing.

  • @muhammadrafiq1720
    @muhammadrafiq1720 ปีที่แล้ว

    There is Ad after each 3 to 4 minets , difficult to concentrate especially with low speed inter et.

  • @theharvi_
    @theharvi_ 9 หลายเดือนก่อน

    ❤thx

  • @AmmarAnjum-h2s
    @AmmarAnjum-h2s ปีที่แล้ว +2

    Why sir you don't talk point to point things..repeating everything again and missing some stuff to talk

  • @mr.patientwolfx5984
    @mr.patientwolfx5984 2 ปีที่แล้ว

    sir what do you think of guvi data science program? can i join.

  • @mohitkumarsingh7318
    @mohitkumarsingh7318 11 หลายเดือนก่อน +1

    Sir pls, also cover SVD , it's a request

  • @viratjanghu945
    @viratjanghu945 ปีที่แล้ว

    Sir please make a video on the independent component analysis and linear discriminant analysis it is my humble request sir please

  • @SohanDeshar-pf6zh
    @SohanDeshar-pf6zh 7 หลายเดือนก่อน

    Good explanation but it might be a good idea to remove one of the "InDepth"s from the video title.

  • @arungireesh686
    @arungireesh686 ปีที่แล้ว

    superb

  • @ramdharavath7542
    @ramdharavath7542 2 ปีที่แล้ว

    Useful

  • @shruti9731
    @shruti9731 10 หลายเดือนก่อน

    ❤❤

  • @somnath1235
    @somnath1235 2 ปีที่แล้ว +1

    What does the covariance and corelation decide ? Does covariance denotes how closely 2 features exist? And does corelation denotes whether the features are directly or inversely proportional?

    • @saisrinivas3066
      @saisrinivas3066 2 ปีที่แล้ว

      covariance only describes the type of relationship whereas correlation describes the type and strength of the relationship between two numerical variables

    • @bhargav1811
      @bhargav1811 2 ปีที่แล้ว

      Correlation is scaled version of covariance !!!!
      Range of covariance = (-inf,+inf)
      Range of correlation = (-1,+1)

    • @Datadynamo
      @Datadynamo 2 ปีที่แล้ว +3

      Covariance is a measure of the joint variability of two random variables. It tells you how two variables are related to each other. A positive covariance means that the variables are positively related, which means that as one variable increases, the other variable also tends to increase. A negative covariance means that the variables are inversely related, which means that as one variable increases, the other variable tends to decrease.
      Correlation is a normalized version of covariance, it gives the measure of the strength of the linear relationship between two variables. It ranges from -1 to 1, where -1 is the perfect negative correlation, 0 is no correlation and 1 is perfect positive correlation. Like covariance, it tells you how two variables are related to each other, but it gives you a more intuitive sense of the strength of the relationship, as it is scaled between -1 and 1.

  • @siddharthmohapatra7297
    @siddharthmohapatra7297 2 ปีที่แล้ว

    Sir I want to ask ...I have no coding skills and background...bcom Background
    Can I do data science masters from pw skills ... everything will be taught from verry basics ???

    • @rutvikchauhan1572
      @rutvikchauhan1572 2 ปีที่แล้ว +1

      You can do it, first learn python , then search data science cources on youtube and on various apps like udemy , coursera , swayam...... And enrolled on it......

    • @siddharthmohapatra7297
      @siddharthmohapatra7297 2 ปีที่แล้ว

      @@rutvikchauhan1572 I have enrolled in pw skills

    • @anuraganand6675
      @anuraganand6675 ปีที่แล้ว

      @Rutvik Chauhan how is you feedback of pw skills data science course?

    • @akindia8519
      @akindia8519 8 หลายเดือนก่อน

      ​@@siddharthmohapatra7297 hi can you please give us feedback of pw skills' data science masters program?

  • @jitendrasahay3847
    @jitendrasahay3847 4 หลายเดือนก่อน

    I have to say : a very short precise material has been elongated irritatingly.
    Repetative statements...

  • @vaibhavyadav-w8g
    @vaibhavyadav-w8g ปีที่แล้ว

  • @shanthan9.
    @shanthan9. ปีที่แล้ว

    Good video but too lengthy

  • @siddhisg
    @siddhisg ปีที่แล้ว +1

    greater than less than symbol though🥲

  • @satyapujari7731
    @satyapujari7731 ปีที่แล้ว

    After every five minutes, there was an advertisement, which made it difficult to concentrate while watching videos.

  • @vishalgupta9620
    @vishalgupta9620 ปีที่แล้ว +1

    noob knows nothing

  • @priyotoshsahaThePowerOf23
    @priyotoshsahaThePowerOf23 ปีที่แล้ว

    BEST