Hello
I’ve noticed that throughout the lectures we are using
m is the number of training examples;
index i from 1 to m to enumerate training examples;
n is the number of features and
index j from 1 to n to enumerate features.
However, in practice labs indexing is reversed: i from 1 to m is used for features, and j from 1 to n is used for training examples.
I am doing the Anomaly detection lab right now, but I’ve seen this in previous labs too - and it confused me every time I see it in a lab. I think using indexing consistently between lectures and labs would make this a little easier. Or am I misreading/misunderstand this completely?
Thank you!
Can you give some screenshots that use m or n differently, but without showing any assignment solution? (because we can’t share learners’ own work here).
In that lab, yes, we used j for samples, and sometimes in the lecture we used i. I can’t say they are always reversed because otherwise I will have to check through the labs one by one.
While we usually use m for samples and n for features, it is, at the end of the day, our job to check the meaning of symbols, because there is no universal standard.
Sounds good. I did run into this issue before (I also posted a question here on the forums). My mistake was - I see index i - and I imagine it to be the i-th training example, not the i-th feature, and from there all my code is wrong
I will pay closer attention in the labs.
Thank you for quick replies, Raymond!
