Nearest Neighbour Model: Use ratings of most similar users Latent Factor Analysis: Solve for underlying factors that drive the ratings
The challenge to recommend items to an entirely new user who has not interacted with any items yet.
Can be used in case of cold-start problem, can recommend to both existing and new users. Uses both collaborative and content based filtering for recommendations.
Ratings and like/not like.
Clicks, watched movies, songs listened to etc.
AUC Score:
Measures ROC AUC metric for the model.
Probability that a random chosen positive example has a higher score than a random chosen negative example.
“People who agreed in the past will agree in the future.”
Based on what a user will like based on the similarity with other users.
Person A likes items 1, 2, 3.
Person B likes items 2, 3, 4.
A should like item 4 and B should like item 1.
“If you like an item you will also like a ‘similar’ item.”
Recommend products which are similar to the ones that a user has liked in the past.