The goal of this project is to build a system that can identify a user's preferences and combine restaurant profiles using better collaborative and content based models. In particular, I implemented K-nearest neighbor and deep neural networks to make recommendations and then compare their performance. The models achieved a good RMSE and MRR score of 1.173 and 0.1315 respectively. Besides the results, I discuss interesting conclusions about the behavior of each model.
Please refer the document yelp_project_report for more details.