Week 2 - Collaborative filtering algorithm question

[Jan_Z]

Link:(https://www.coursera.org/learn/unsupervised-learning-recommenders-reinforcement-learning/lecture/ppY77/collaborative-filtering-algorithm)

I am confused here about the first regularization term in the ‘put them together’ part; Can someone help explain why the ‘w’ term has a subscript k on it? I thought k just specifies between the different features (romance (1) or action (2))? Thank you!

[rmwkwok]

Hello @Jan_Z,

Thank you for sharing the link and the screenshot!

You are right that a k means a feature, but for each feature, shouldn’t there be a w and a x at the same time, otherwise, what would x multiply to?

Cheers,
Raymond

[Jan_Z]

Right, and this is highlighted here:

So when k goes from 1 to n, is n equal to 2 in this case. Would k=1 and 2 for each weight w just specify which weight is selected inside of the weight vector?

[rmwkwok]

Hello, @Jan_Z,

Yes, I agree with you!

FYI:

Cheers,
Raymond

[Jan_Z]

Thanks!