This code gives a brief understanding of how to use the surprise library for RecSys. Dataset used is MovieLens-100k.
MovieLens-100K : This data set consists of:
-100,000 ratings (1-5) from 943 users on 1682 movies.
-Each user has rated at least 20 movies.
The flow of the project is:
1-EDA
2-Model Selection
3-Tuning algorithm parameters
4-Training and Testing the model
5-Analysis on the predictions
6-K Recommendations
-
Finding the threshold value using the F1 score
-
Finding the optimal value for K using precision and recall
7-Recommendations to users
8-Comparing predictions with the user's history.