At 3.03 , xi ' should be ( (ui) dotproduct (xi) )/ mod(ui) ,however it shows mod (ui) ^2 as told in projection and unit vector lecture at 2:24 link- th-cam.com/video/fbrMJbMcGoA/w-d-xo.html
We standardise each feature or axis of data. When we find the maximum variance direction, that need not be axis parallel. Oftentimes, it is a linear combination of other features and hence is a direction which is not parallel to any axis
![เปิด อาณาจักร หมื่นล้าน "สาระตั้ม" l [Nickynachat]](http://i.ytimg.com/vi/3XE0nPF4YkQ/mqdefault.jpg)
Projection is a vector not a scaler right..... I think there's a slip of tongue...
Beautiful lecture:)
3:09 you say "we learned" what video do you mean?
How xi can be (nx1) it should be (nxd) as you project point on to u1
what is col standerization, , how is it
equal to 0
(x-meanx)/sigma I guess
basically mean is equal to zero
At 3.03 , xi ' should be ( (ui) dotproduct (xi) )/ mod(ui) ,however it shows mod (ui) ^2
as told in projection and unit vector lecture at 2:24
link-
th-cam.com/video/fbrMJbMcGoA/w-d-xo.html
After column standardization variance will also become 1 how can we maximize variance term?
We standardise each feature or axis of data. When we find the maximum variance direction, that need not be axis parallel. Oftentimes, it is a linear combination of other features and hence is a direction which is not parallel to any axis
Why do you maximise variance?
to get maximum information..
Also variance refers to variability. By maximising it you get a diverse data.