AI meets Big Data: User Analytics and Personalized Recommendation Based on Location Data
CHAN Gary Shueng Han
User locations can be obtained through many means, e.g., GPS, apps, or people sensing. These data is inherently noisy, sparse and irregular. In this project, you will study and implement how to combine AI (Artificial Intelligence) machine learning techniques with big data (on user locations) to automate the following: data cleansing, trajectory inference, behavior mining, and prediction/recommendation.
UROP1100 UROP2100 UROP3100 UROP4100
Students will work in a rigorous R&D team setting to propose, study, implement and experiment novel algorithms. Machine learning, programming and computational skills will be involved.
Applicant's Learning Objectives:
1) Achieve knowledge on how to use AI techniques to extract or collect large-scale data;
2) Achieve knowledge on how to use statistics and optimization to automate data cleansing and denoising;
3) Achieve knowledge on how to mine user behavior out of the cleansed data;
4) Achieve knowledge on how to make predictions and recommendations given user behavior