Basic splitting of the user-item interaction matrix into train and testing part.
Useful for when data doesn't have time dimension.
Usually during model tuning it worth to keep some
x matrix as hold-out data set.
x could be splitted in 2 parts - train and test.
Model tries to predict test data using train
train_test_split(x, test_proportion = 0.5)
sparse user-item interation matrix. Internally
- proportion of the observations for each user to keep as "test" data.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.