My name is Sung-En Chang. I am currently a Ph.D. student at Northeastern University, working with Ian En-Hsu Yen and Rose Yu on Learning Tensor Latent Features for Hyperspectral image processing. Before that, I am also a post-bachelor student at the Computer Science School of Tamkang University and I received my B.S./M.S. from the Finacial and Economic Law departments of Fu Jen Catholic University. My research focuses on Large-Scale Machine Learning, Optimization and their applications. Please see my CV for further information.
- Ph.D. in College of Computer and Information Science, 2018 - 2023 (expected).
- B.S. in Department of Computer Science and Information Engineering, 2016 - 2018.
Fu Jen Catholic University
- M.S. student, Department Financial and Economic Law, 2011 - 2015
Fu Jen Catholic University
- B.S. in Financial and Economic Law, 2007 - 2011
Working ExperienceNational Taiwan University
Research Assistant, Mar. 2015 - 2018.
Develop Max-Cut / SDP solver for solving convex relaxation of Binary Latent-Variable Models. (with Ian E.H. Yen, Wei-Cheng Lee and Shou-De Lin).
Intern Data Scientist, July. 2016 - Feb. 2017.
Network packet anomaly detection (with Jen-Lin Lai and Shou-De Lin).
- Co-Founder, Jan. 2014 - June. 2016.
Turing Trademark Attorny, a trademark search engine and registration website (with Shis-Wen Huang and Ian E.H. Yen)
- Lawyer, Jan. 2014 - Dec. 2014.
Civil law, Crime law, orporation law, Securities regulation and Insurance law
Research Assistant, Oct. 2012 - Jun. 2013.
The Taiwanese legal research method in cyberspace(with Ching-Kuen Ueng).
Learning Tensor Latent Features
Sung-En Chang, Xun Zheng, Ian E.H. Yen, Pradeep Ravikumar and Rose Yu. (under review) 2018.
- Generalized Symmetric Matrix Factorization for Overlapping Community Detection Xun Zheng, Sung-En Chang, Venu Satuluri and Eric P. Xing (under review), 2017.
MixLasso: Generalized Mixed Regression via Convex Atomic-Norm Regularization.
Ian E.H. Yen, Wei-Cheng Li, Sung-En Chang, Shou-De Lin and Pradeep Ravikumar. (under review), 2017.
2017Latent Feature Lasso [pdf] [slide] [poster] [talk]
Ian E.H. Yen, Wei-Cheng Li, Sung-En Chang, Arun S. Suggala, Shou-De Lin and Pradeep Ravikumar. In International Conference on Machine Learning (ICML), 2017.
A Boolean (0/1) Quadratic Programming solver employing the algorithm proposed in Mixing Method for solving large-scale SDP, which can handle problems of millions of variables.
A solver for anomaly detection, which can classify and predict next type of event with high accuracy.
A website for trademarks, which can help users register their trademarks and raise the acceptability.
Familiar: Python, Matlab