Commercial entities, such as search engines, advertisers, media platforms (e.g., Netflix, Amazon), and financial institutions, leverage recommender systems to recommend content, predict customer behavior, ensure compliance, and assess risk. This course offers a comprehensive overview of predictive models for recommender systems, encompassing content-based and collaborative filtering algorithms, matrix factorization, and deep learning models. Additionally, the course provides hands-on experience with implementing existing recommender systems using Python.

👌 What you’ll learn:

  • Gain insight into the fundamental principles underlying various recommender systems approaches, including correlation-based collaborative filtering, latent factor models, and neural recommender systems.
  • Develop hands-on experience implementing and analyzing recommender systems for real-world applications using Python, sklearn, and TensorFlow.
  • Learn to select and design suitable models tailored to specific applications.

🏗️ Prerequisites:

  • Calculus & Linear algebra: inner products, matrix-vector products, linear regression (OLS).
  • Basic Statistics: Basics of distributions, probabilities, mean, standard deviation, and other fundamental concepts.
  • Python: Familiarity with basic Python syntax and experience with Numpy, pandas, and TensorFlow libraries.
  • (Recommended) Complete a Machine Learning Crash Course (in-person, online, or self-study) or possess equivalent knowledge
  • ⏲️ Lectures: Thu 12:30PM - 3:15PM
  • 🎒 Lecture/Recitation Location: Mong Man Wai Bldg 710
  • 💻 HW submission: BlackBoard
  • ⌨️ colab: notebook or click Open in Colab

Homepage GitHub Open In Colab

All students welcome: we are happy to have audiences in our lecture.

📋 Reference Textbooks

The following textbooks are relevant resources, although none of them perfectly align with the scope and content of our course.