Using the mixture model for collaborative filtering

$$ Bookmark this page

Project due Aug $21,202417$:$29$ IST

Reporting log likelihood values on Netflix data

$0\mathrm{.}0/1\mathrm{.}0$ point $($graded$)$

Now, run the EM algorithm on the incomplete data matrix from Netflix ratings netflix$\_$incomplete.txt $.$ As

before, please use seeds from $0,1,2,3,4$ and report the best log likelihood you achieve with $K=1$ and

$K=12.$

This may take on the order of a couple minutes for $K=12.$

Report the maximum likelihood for each $K$ using seeds $0,1,2,3,4$ :

Log$-$likelihood ${|}_K=1=$

Log$-$likelihood ${|}_K=12=|$