I am (CV, Linkedin) a Ph.D. candidate in Industrial & Systems Engineering at Lehigh University, working with Prof. Katya Scheinberg on developing practical algorithms (and their theoretical analysis) for machine learning problems and many other real-world applications.
Prior to study at Lehigh, I earned my undergraduate degree in Control Science and Engineering from Chu KoChen Honors College in Zhejiang University, Hangzhou, China. During summer I worked at industrial research labs on projects in different fields such as ads targeting, machine learning, hierarchical classifications, time series forecasting, traffic control, etc.
An efficient algorithm for optimizing large-scale regularized functions. Current C/C++ (with MATLAB interface) implementations can handle
The algorithm/implementation can be easily extended to deal with smooth loss functions other than least square and logistic loss. See for yourself.
Efficient Hierarchical Multi-label Learning, submitted, 2014
Practical Inexact Proximal Quasi-Newton Method with Global Complexity Analysis, submitted to Mathematical Programming Series A, 2013
HIPAD - A Hybrid Interior-Point Alternating Direction algorithm for knowledge based SVM and feature selection, Proc. of Learning and Intelligent Optimization Conference, 2014
A Fast Decomposition Approach For Transportation Network Optimization, Proc. 19th IFAC World Congress, 2013
HiClass - An efficient hierarchical classification system, Siemens Technical Report, 2011
The Application of Frequency Family Separation Method in Rolling Bearing Fault diagnosis Based on Empirical Mode Decomposition, Proceedings of the 29th Control Conference (CCC), China, 2010.
Hierarchical Multi-label Ads Targeting, Yahoo Labs, Sunnyvale, CA, 2014.
Efficient Quasi-Newton Proximal Method for Large Scale Sparse Optimization, NIPS Optimization for Machine Learning Workshop, 2013; Informs Optimization Conference, Houston, 2014; SIAM Conference on Optimization, San Diego, 2014; SIAM Annual Meeting, Chicago, 2014.
Practical Inexact Proximal Quasi-Newton Method with Global Complexity Analysis, Informs Annual Meeting, San Francisco, 2014.
A simulation-optimization approach for traffic control re-optimization in the presence of incidents, IBM T.J.Watson Research Center, Yorktown Heights, NY, USA, August 2013.
Efficiently Using Second Order Information in Large l1 Regularization Problems, MOPTA Conference, Bethlehem, PA, USA, August 2012.
Experiments with model-based algorithms for Derivative-Free Optimization, COR@L Seminar Series, Lehigh University, Bethlehem, PA, USA, 2010.