brick-mover / w18-ee219-data-mining

UCLA W18 EE219 Projects

EE219

Project 3: Collabrative Filtering

• Build a recommendation system to predict the ratings of the movies in the MovieLens dataset.
• implement and analyze the performance of two types of methods:
  • Model-based collaberative filtering
  • Neighborhood-based collaborative filtering

Project 4: Regression Analysis

Project 5: Popularity Prediction on Twitter

• knowing current (and previous) tweet activity for a hashtag, can we predict its tweet activity in the future?
• The available Twitter data is collected by querying popular hashtags related to the 2015 Super Bowl spanning a period starting from 2 weeks before the game to a week after the game.
• We will use data from some of the related hashtags to train a regression model and then use the model to make predictions for other hashtags.
• Design custom topics