Hi Neerja,

Thanks for the article! I am building a recommender system too and my model has very high AUC (0.95+)but very low precision at 5 (~0.1)and recall at 5 (~0.1). It seems that my classifier achieves good performance on the positive class (high AUC) at the cost of a high false negatives rate (or a low number of true negative). I used 10% of the whole data as the test set. Have you ever run into problems like this and can you give me some advice on moving forward with model tuning?

Thanks, Tina