Description
Coding Assignment 4: Recommender System
Build a movie recommender system based on the MovieLens 1M Dataset.
Source
Download the MovieLens 1M Dataset. You’ll find four files: README, movies.dat, ratings.dat, and users.dat. Check the readme file to understand the format of the other three files.
What you need to do?
Assume your working directory contains the following files: movies.dat and ratings.dat that are exactly the same as the ones from the MovieLens 1M dataset;
Step 1: Set the seed at the beginning of your code to be the last 4-dig of your University ID.
Step 2: Create
- train data that contains about 60% rows of the ratings.dat from the MovieLens 1M dataset (of the same format);
- test data that contains about 20% of the user-movie pairs from the ratings.dat from the MovieLens 1M dataset.
Note that the train data contains just 60% of the original ratings, so it is possible some movies in movies.dat or users in users.dat do not appear in the training, but in test.
Step 3: Build two models to predict the movie rating for each user-movie pair in the test data.
Step 4: Report the RMSE (Root-mean-square error) of your two models on the test data.
