Scholar
Fan-Keng Sun
Google Scholar ID: sfEwE4gAAAAJ
Ph.D. student at EECS, MIT
Citations & Impact
Citations
1,640
H-index
11
i10-index
12
Publications
20
Co-authors
0
Contact
Publications
Resume (English only)
Academic Achievements
- Proposed to use self-generated data to enable lifelong learning for GPT-2; as an undergraduate team, won top-3 in the ISPD (physical design) Contest.
Research Experience
- During his Ph.D., his research interest is foundation models for long-term sequence modeling. He has worked on various topics of deep learning for time series, including specialized small models and general foundation models, from anomaly detection to forecasting, and from generalization to drift adaptation. Throughout his Ph.D., he has been heavily involved in long-term industry collaborations, such as anomaly detection for semiconductor manufacturing with Analog Devices and drift adaptation in machine health with Lam Research.
Background
- An avid entrepreneur who bridges academic theory and real-world practice, with experience in co-founding several startups, securing spots in top accelerator programs, and attending structured MBA courses at MIT/Harvard.
Miscellany
- Loves to explore the world, which led him to participate in the International Geography Olympiad (iGeo) and represent Taiwan to win a Silver medal. (Bonus fun fact: that’s how he met his wife!)
Co-authors