Fashion Product Rating Prediciton for Retail Website Datasets

Originally an Assignement for CSE158 - Recommendation System

Highlights of This Project

• Included collaborative filtering feature with user and item interaciton to improve performance
• Included item similairty feature using Jaccard and/or Cosine distance
• Tested the performance of Random Forest, KNN, a couple of popular regressor on the dataset
• Improves the model performance by dividing the single-multiclass prediction task into 2 different task for better performance
• Include attempted latent factor approach's implementation in Tensorflow