About meI am joining TU Delft as an Associate Professor in fall 2018, after a decade of industry experience at the IBM Research Zurich Lab. My research interests lie in the distinct areas of cloud computing, big data, machine learning, and system dependability. I received my PhD from the Penn State University, and I completed my undergraduate studies at National Taiwan University and British Columbia University.
I am recruiting motivated students and postdocs to work on private machine learning, adversarial learning, fair learning, and their applications. Please send me your CV and research interests.
Over the years, I have worked on resource management problems of various computing systems, such as web services, cloud datacenters, and big data processing systems. The central theme of my research is to strike an optimal trade-off between the application performance, e.g., job tail latency and system dependability, and resource allocation. I draw methodologies from analytical models, machine learning, and develop system prototypes for experimental validation.
My recent focus on big data analytics and processing systems leads me to address exciting questions, such as:
- (i) slim private learning: how to maximize the discovery capability of (deep) machine learning algorithms, while maintaining data privacy at a minimum amount of resources
- (ii) robust and active learning: how to make learning algorithms robust against adversaries that maliciously manipulate the data input
- Keynote talk at IEEE International Symposium on Parallel and Distributed Computing 2019.
- Keynote talk at DCPerf workshop, in conjunction with INFOCOM 2019.
- Our paper "Robust Anomaly Detection on Unreliable Data" has been accepted at IEEE/IFIP Conference on Dependable Systems and Networks (DSN) 2019.
- Our paper "Duo Learning for Classifications with Noisy Labels" has been accepted at Continual Learning Workshop, in conjunction with Neural Information Processing Systems (NIPS) 2018.
- Our paper "Automatic Privacy and Utility Preservation for Mobility Data: A Nonlinear Model-Based Approach" has been accepted for publication in IEEE Transactions on Dependable and Secure Computing.
- Our paper "Model-Driven Computational Sprinting" has been accepted at Eurosys 2018.
- I am currently involved in the following conference organizations and journal editorial boards:
- Editorial boards:
Associate editor at IEEE Transactions on Service Computing,
Associate editor at IEEE Transactions on Network and Service Management,
Associate editor at IEEE Cloud Magazine,
Guest editor at ACM Transactions on Autonomous and Adaptive Systems.
- Conference organizers:
2019 ICAC program co-chair http://icac2019.cs.umu.se/,
2018 Middleware industry track co-chair
- Technical program committees:
INFOCOM 19, IPDPS 19, DSN 19, CCgrid 19
AwardsDelft Technology Fellowship, 2018,
ACM ICAC Best Paper Award nomination, 2017
ACM ICAC Best Paper Award nomination, 2016
ACM eEnergy Runner-up Best Paper Award, 2015
IEEE/ACM CCGrid Runner-up Best Paper Award, 2015
IBM Outstanding Scientific Achievement Award, 2014
IEEE/IFIP DSN Best Paper Award nomination, 2014
IBM All-level Scientific Achievement Award, 2012
IEEE HPDC Best Presentation Award nomination, 2012