Lecture 20- Linear Discriminant Analysis ( LDA) (with Solved Example)
Vložit
- čas přidán 5. 09. 2024
- This video is about Linear Discriminant Analysis.
If you are interested in building cool Natural Language Processing (NLP) Apps , access our NLP APIs at www.firstlangu... . Also for NLP product development and consultation, please reach out to us at info@firstlanguage.in
I searched a lot for LDA example but finally, I found the mathematics behind the calculation behind.
Very useful video. Thank you maam. Appreciate your hard work
so true my 1 h search ended with this !
Thank you Madam. This is plain old school teaching at it's best. No fuss, just imparting knowledge !
You have calculated the Sw^-1 correctly later but used the wrong values to find the vector.
Yes😂
@@parvathykrishna.r536 we wasted 20 minutes of our life give a disclaimer next time
Hello Maam, at 9:17 HOW you calculated the value [-0.91 -0.39] ^T. would you elobrate this maam
So far the best LDA explanation I have found!
Finally found a detailed explanation of LDA. Thank you Maam.
This is the best video I have found on the subject. Great work!
Yes, is wrong. very confusing.
Can you explain why the inverse you used before and after is different?
In Sw inverse you have substituted the another value mam, if we try we are getting different answer
The explanation was very detailed and nice mam. In calculation of s1, the fourth element should be 2.37. It is written as 2.69. I think it was a calculation mistake. Please check.
Really helpful. Thanks. but the inverse Invert of SW is not matching when we calculate it. But the concept is clear
bro which one is the correct answer?
Mam it was a Very beautiful step by step explanation. Finally understood the steps
best lectures on ML Thank u so much maam for u r efforts
very Nice Explanation,Mam.Needs More tutorial From You.
Oh wow, I actually get something for once, thank you so much!
Best math explanation on LDA. Thank you !!
Thank you ma'am! Loved the patient step by step approach. I just can't relate the notions of maximizing/minimizing an expression and finding the eigenvector. For example in LDA we are trying to maximize the between group variance and minimizing the within group variance, then how does that relate to finding the eigenvector to the quotient of the two covariance matrices?
Clearly explained mam! Very easy to understand the concept! Thank you so much
from 8.19 i get confused.. to my understanding (Sw)^(-1) calculated in the note part.. so we need to calculate that value along with SB matrix for getting v1 and v2.... in another method also you are not mentioning (Sw)^(-1) value again changing the diagonal and sign values... please do explain clearly mam
did u get answer?
excellent explanation madam...i am cbit hyderabad ece faculty thank you madam
a way better explanation than my profs does it!
for computing SB ,u have written (m1-m2) *(m1-m2)transpose but u have done (m1-m2)T *(m1-m2) which is wrong.please explain
You deserve more
WoW! Good Explanation. Thank you.
Best and the simplest explanation of LDA
Thankyou mam.. I have a small query, Please could you mention the X-axis and Y-axis.
Very informative video madam. Can you please suggest any book having all the mathematical intuition behind the ML algorithms?
thank you so much maam, I finally understand it
when u calculate the covariance matrix it should be divide by n-k k which is no of classes
2 in this case but i am not sure about this part
Thank you mam, it is absolutely great session. But a small request, if you explain same thing by taking attributes it would be great for us
Nice explanation madam. Thank you so much. The inverse written at 8:59 and used before this time is wrong. After wards its correct. Please check.
Yes, is wrong. very confusing.
Can you explain why the inverse you used before and after is different?
At 7:00 , the ų1-ų2 is incorrect, it should be [-5.4 1]
Thanks from Maharashtra
amazing explanation
fisher discriminant analysis and Linear discriminant analysis are same?
yes
7:35 how did you made that matrix? I don't get it 😭, can someone plz explain quickly, I have exam tomorrow 😱😱
Nice explanation. Can you suggest any text book for pattern Recognition having good amount of numerical
Maam Your covariance matrix for class 1 is wrong. Please recheck the calculation.Because when we calculate covariance we divide the value by n-1 not n
Hello Maam,
could it be possible to explain further LDA to extend to liner polynomial Kernel trick with the same numerical example
Thank you very much!
Dear Sir and everyone, I have a question that need help now. Could you show me how to calculate MSE in LDA ?? Thank you
Thank you so much for lecture. it was clear and easy to understand
mam you have amistake in inverse formula
Excellant video mam
Which textbook you are using. Kindly tell
Helpful ❤❤
Thank you so much ma'am
Thank you so much
thank you.
thank you so much mam.
Mam... how to find 3 class feature???
Use 3 mean values!!!!
Why your Sw^-1 is different in the final formula and computation?
she has taken wrong sw-1 in first formula
Can you please link a PDF of your notes in this video?
thanks alot madam!!!!!!
This video was extremely useful. Thank you so much!
S is not covariance but rather scatter
nice explation mam luv you
thanks you !
Dear ma'am,
At 5:55 you are dividing each of the entries by 5 (total number of samples under the class) after summing values from the 5 matrices(for the 5 individual samples), but this division step is not present in the formula at 4:56 which only shows the multiplication and the summation parts? Is this a typo or am I missing out something here? Thanks in advance.
1/N should come bro in that formula part...
Understood complete problem. Thanks
Hello Dr. Subalalitha,
After following one of your videos on LDA analysis,
1. is it okay to use the global mean of the matrix while calculating the individual class covariance or, we should use the group mean (just as you did in your video
2. How do i draw the dimension reduction graph (y=W^T * X)
permission to learn mam
Ty aunty
madam matrix S2 is not correct
thanks for informing but I cross checked it seems to be correct sir .
@@subalalithacn5827 is it okay now?
Too many things going wrong here. (Explanation is good, but solving the equations)
Terribly wrong calculations for God's sake!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
in S_1 2.6 should be 2.64
Hi, remember in academics it's always important to give appropriate credits, please mention that you took the examples from the IIT-kharagpur's ppt from the link www.facweb.iitkgp.ac.in/~sudeshna/courses/ml08/lda.pdf. Thanks
Hello Maam, at 9:17 HOW you calculated the value [-0.91 -0.39] ^T. would you elobrate this maam
I am also confused