Hi! In chemometrics we normally work by arranging our matrices in such a way that samples are in the row direction and variables are in the column direction. That is, if we have a dataset of 5 samples and 20 discrete variables, our matrix X(I,J) will have dimensions (5,20). In your case, 20 discrete variables means J = 20 Cheers Q&T
I understood the concept of PLSR but the algorithm is something really difficult to digest. Do you have a vedio that explains the algorithm somehow in easy language? But really thank you for these videos. you have no idea how much helpful they are to me!
This is a great explanation, but I am still confused by what indicies I and J represent - if I want to use this in optical spectroscopy and I sample a spectrum of a range of different compounds at 20 discrete wavelengths, does this mean I = 20 or is J = 20? thanks!!
Thank very much for this presentation. For the plotting of X and Y loadings, are we plotting the first component of Y loadings versus first component of Y loadings?
This is great. One thing though. I'm not sure how the Score of X for Remus was so close to 0? It seems like the turquoise line connecting the loading vector to remmux is pretty large
The length is not related to the score. If you read the variable score in a coordinate axis you check the actual axis, not the length of the line going to the axis.
It's incredible! This is the most useful lecture about PLS I've experienced. The video has fulfilled all my confusions about PLS. Thank you very much.
Excellent video series! I loved the interpretation of loading scores in the PCA series as well!
Thanks so very much. This is so very nice for us to hear. It is very rewarding to know that the videos are helpful. Thanks
me too, thx so much. I am so depressed reading these sort of stuff for a whole and never get a clue until I reach this video.
Hi!
In chemometrics we normally work by arranging our matrices in such a way that samples are in the row direction and variables are in the column direction. That is, if we have a dataset of 5 samples and 20 discrete variables, our matrix X(I,J) will have dimensions (5,20). In your case, 20 discrete variables means J = 20
Cheers
Q&T
Thank YOU, indeed! Thanks for the positive comment and keep watching videos and whenever you need, just write to us!
Cheers
Q&T
Thank you this is fantastically explained
Thank you, very clear explanation.
Thank you! Helped me very much with my bachelor thesis!
great video
Very very helpful man.
I understood the concept of PLSR but the algorithm is something really difficult to digest. Do you have a vedio that explains the algorithm somehow in easy language? But really thank you for these videos. you have no idea how much helpful they are to me!
This is a great explanation, but I am still confused by what indicies I and J represent - if I want to use this in optical spectroscopy and I sample a spectrum of a range of different compounds at 20 discrete wavelengths, does this mean I = 20 or is J = 20?
thanks!!
Thanks ever so much
Thank very much for this presentation. For the plotting of X and Y loadings, are we plotting the first component of Y loadings
versus first component of Y loadings?
Normally, it would be the scores
This is great. One thing though. I'm not sure how the Score of X for Remus was so close to 0? It seems like the turquoise line connecting the loading vector to remmux is pretty large
The length is not related to the score. If you read the variable score in a coordinate axis you check the actual axis, not the length of the line going to the axis.
thank you soooo much
Hi.I want to ask , can i use PSL when I do PCA only on X's in multivariate regression??
+waed abead You can use PLS any time you want to do regression.