Skip to main content
Open Access Publications from the University of California


UCLA Electronic Theses and Dissertations bannerUCLA

Application of machine learning techniques to recommendation system

  • Author(s): Gao, Mei
  • Advisor(s): Wu, Yingnian
  • et al.

In this paper, I applied several machine learning techniques, including Latent Dirichlet allocation (LDA), deep Convolutional Neural Networks (CNN) as well as Ranking SVM to build a restaurant recommendation system using the Yelp Dataset. The idea is that, compared to algorithms that are previously commonly used in recommendation system, i.e. Pearson similarity and clustering algorithms, the application of machine learning techniques such as LDA, CNN and SVM to recommendation has been a new area and not systematically studied yet. Topic models such as LDA could allow us to learn the latent subtopics in review texts, which then can be used in personalized recommendation. Deep CNN could be used to extract the underlying features from food images and combined with Ranking SVM to provide users with the best food image of a restaurant. In this paper, I made several attempts to introduce these machine learning techniques to the construction of a restaurant recommender engine and results have shown that the efficiency of the recommendation system could be largely improved with these techniques.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View