Enhance Ads products with AI/ML technology

Cloud platform, shared services, recommender system.

Data Mining, Machine Learning, Trajectory Data, Spatial Temporal Data.

Current Project:

Interpretable Temporal Mining for Contrastive Driving Behaviors:

Proposing new models to mine the contrastive driving behaviors generated by participants diagnosed with ADHD before and after taking their ADHD medication. Moreover, developing ADMM-based optimization algorithms and implementing solvers in Python to find the optimal solution of the proposed models.

Past Projects:

1. A Uniform Representation forTrajectory Learning Tasks’http://doi.acm.org/10.1145/3139958.3140017

2. TrajViz: A Tool for Visualizing Patterns and Anomalies in Trajectory

\par Developing algorithms and visualization Java application for mining and visualizing trajectory patterns and anomalies.

https://github.com/qingzheli/TrajViz

Working on multiple software engineering projects:

1. Data preprocessing, data augmentation, training and evaluating deep learning models on remote Linux servers.

2. Data preprocessing, designing and implementing novel deep learning architecture from scratch, design and implementing comprehensive experiments on training and evaluating the proposed model.

3. Building Restful web service to get data from Blackboard server, compile the data to the schema in EAB system, and post the compiled data to EAB server.

4. Processing 40 TB MRI data in ConnectomeDB to generate the structural and functional connectivity datasets for deep graph translation models.

5. developing java application to mine and visualize frequent and anomalous trajectories on OpenStreeMap