dannydhkim/movierating

Movie Rating Prediction

Read the full report here

Project Description

With the onset of the big data age, we are flooded with a large ocean of information that we cannot sort through. Recommendation systems are the basis for navigating much of the data we have today from Amazon's "Frequently Bought Together" recommendations to the recommendations as seen in Netflix. In Stanford's CS 102 class, we were expected to use the MovieLens ratings data, collected by GroupLens Research, to predict how users will rate movies they haven't watched based on their past movie ratings.

Algorithms

This project implements the KNN and FunkSVD algorithms

Evaluation

These algorithms were evaluated using RMSE and MAE

File Description

MovieRating.pdf - Full report on the project

EDA.html - exploratory data analysis

Dataset - Given MovieLens dataset

src - python files and jupyter notebook

Predictions - Results from algorithm implementations