I'm a fifth-year PhD student in Computer Science Department of Carnegie Mellon University, working with Prof. Jeff Schneider. Before that, I received my Bachelor of Engineering from Department of Computer Science and Technology in Tsinghua University, China.
Machine Learning: Transfer Learning, Active Learning, Kernel Methods.
I'm specially interested in transfer learning with model shift (conditional distribution change), for regression/classification problems or learning shifting distributions. Applications include automated yield estimation based on images, and distribution estimation of locations on traffic data.
Nonparametric Risk and Stability Analysis for Multi-Task Learning Problems
A general framework for multi-task kernel learning with stability analysis, and connection with transfer learning algorithms
Generalization Bounds for Transfer Learning under Model Shift
Generalization error bounds for transfer learning under model shift
Active Transfer Learning under Model Shift
Proposed transfer learning algorithms that allow changes in conditional distributions
Flexible Transfer Learning under Support and Model Shift
Transfer learning with flexible transformation on both features and labels
Active Search on Graphs
a soft-label model and the impact criterion for active search on large graph datasets
An Impact Criterion for Active Graph Search
proposed an impact criterion for active graph search