Matrix Factorization

Decomposes the user-item matrix into two smaller matrices (user-taste and item-taste). Each user and item gets a short vector of latent factors, hidden taste dimensions found from rating patterns, and their product fills the gaps.

When to use

The lecture presents it as a glimpse beyond the course: "PCA for recommendations", and a later-phase option for a growing platform. The factors are not labelled by hand.

See Lec 6.

Appears in

No references yet.