$14$ marks$)$ Rating prediction is an important task for a recommender system. Try to implement a

recommendation model using the Goodbooks$-10$k dataset to predict user ratings. You have two options

based on your available compute power:

$($a$)$ Option $1($Full Dataset$)$: For students with sufficient compute power, use the entire dataset $(6$

million ratings for $10,000$ books by $53,000$ users$).$

$($b$)$ Option $2($Limited Dataset$)$: For students with limited compute power, use a subset of the data.

Limit the dataset to the top $5,000$ users and top $3,000$ books based on the number of interactions

$($ratings$).$ In your submission, clearly state that you are using the subset due to computational

constraints.

$($c$)$ You can implement the model using user$-$based collaborative filtering, item$-$based collaborative

filtering, or matrix factorization.

$($d$)$ Split your chosen dataset $($full or limited$)$ into a training set and a test set.

$($e$)$ Display the Mean Absolute Error $($MAE$)$ and Root Mean Squared Error $($RMSE$)$ on the test set.