/movielens-analysis

Analysis and model implementations for movie recommender systems around the MovieLens dataset

Primary LanguageJupyter Notebook

Collaborative Filtering Applied to MovieLens Data

This repo contains my analysis of the MovieLens 100K dataset with implementations of various collaborative filtering algorithms, including similarity-based methods and matrix factorization methods using Alternating Least Squares (ALS) and Stochastic Gradient Descent (SGD). My analysis is separated into several notebooks. Click on the links below to view each one.

  1. Exploratory Data Analysis
  2. Baseline Models
  3. Similarity-Based Collaborative Filtering
  4. Alternating Least Squares (ALS)
  5. Stochastic Gradient Descent (SGD)
  6. Comparisons Between Models
  7. Recommender Notebook: This notebook lets you choose a user in system and get the top 3 recommendations for them using your choice of the best user-based, item-based, ALS, or SGD model.