at 16:45 bhaiya said that for a matrix of (5 x 300) we will be taking the mean across the rows to get the average vector, but shouldn’t the mean be taken across the columns to get an average vector of shape 300. I know in video the shape comes out to be 300 , but doesn’t it implies the the vector would have been of the shape (300 x 5), Its a silly confusion that i am facing, plz if u can clarify a little? infact at 16:51 bhaiya highlight columns for which we are taking average for …
A little doubt in video concept
Hi @saksham_thukral,
You have a very natural doubt and quite common too. The terminology used is correct. The mean is a row-wise mean. Let me explain how. The shape of the vector is 5*300 i.e. it has 5 rows. Whenever we take a row-wise mean, it means that multiple rows would be combined (averaged) to make a single row i.e. in this example 5 rows would be combined to make a single (1, 300) vector. Here, we can discard the singleton dimension and our output vector becomes (300,).
(Similarly a column-wise mean would have meant (5, 300) -> (5, 1), multiple columns combined).
Hope this helps!
I hope I’ve cleared your doubt. I ask you to please rate your experience here
Your feedback is very important. It helps us improve our platform and hence provide you
the learning experience you deserve.
On the off chance, you still have some questions or not find the answers satisfactory, you may reopen
the doubt.