Score decreases when features normalization enabled

Nobody has claimed this yet.

Hi,

I train my LightFM model with no features for both users and products (I create a pure interaction matrix) based on implicit interactions. I score my model using mean average precision. In one of my experiments, I tried to add one feature for every user (one-hot encoded, so I added ~10 columns to the user features matrix - I built it using `Dataset` class). Unfortunately, after this operation, my score decreased about 2 times. Interestingly, when I set `normalize=False` in `build_user_features` method, the score returned to the previous value (approximately).

Am I doing something wrong or is it a well-known effect that the score drastically decreases after normalization? Why?

Parameters for my model: `no_components=150`, `learning_rate=0.06`, `loss='warp'`. I trained the model with ~750 000 users and ~200 000 products.

Start with Dataset.build_user_features and reproduce the comparison between normalize=True and normalize=False using the reported one-hot user features, WARP settings, and MAP evaluation. Done means determining whether normalization explains the score change or exposes a defect, then documenting the finding or correcting the behavior if it is erroneous.

Written by the indexing model from the issue text.