• View
• Print
• Recent Changes
• Login

Collaborative Filtering lab: filtering the Netflix dataset

Download the dataset from the original Netflix Prize competition:
http://archive.ics.uci.edu/ml/datasets/Netflix+Prize

The goal of the original contest was to predict the value of missing ratings from the training set.
The correct answers are now available in the grand_prize.tar.gz file.
Use the training data from nf_prize_dataset.tar.gz to train your own system and make recommendations based on what you learn.
Feel free to tackle the original goal of the contest and to try out some of the successful algorithms.

Reference papers on the contest can be found online and are linked to from the data set website at the bottom of the page.

This lab can be extended into your group project for this unit if you are interested. This lab could also form the basis of a class-wide project if that is of interest.