Best Neighbors Learner-Based Algorithm

Rating Prediction

The purpose of a CF algorithm in this context is to recommend to the learner a list of unrated contents sorted by decreasing value. Subsequently, the model will estimate the unknown rating for the contents (books) that the learner's neighbors have rated but the target learner has not. The rating estimation is computed using the following formula:

	Where wij represents the similarity between the content i and one its neighbors j.

Here is the algorithm