An idea that is not dangerous is unworthy of being called an idea at all.
I am primarily interested in fundamental research questions in statistical machine learning. My goal is to invent new statistical models, inference methods and computational algorithms.
Statistical machine learning is a vast field that has rapidly changing the landscape of Artificial Intelligence (AI). My collaborations and I have worked on many subareas in machine learning. Our past (and ongoing) research topics include: (1) unsupervised learning including probabilistic latent variable models and dimensionality reduction; (2) supervised learning especially under data paucity (or learning with small data): multi-task learning, transfer learning, zero-shot learning and domain adaptation; (3) representation learning for automatically inferring useful features from data: learning kernels and metrics, deep learning architectures, etc; (4) large-scale machine learning systems and algorithms: distributed optimization, large-scale kernel methods, etc.
My research agenda is often inspired by many AI problems: speech and language processing, computer vision, robotics, and others.
Most recently, I have branched into applying machine learning (and other AI techniques) to life and medical sciences. Many unique opportunities and challenges arise in those domains. For example, I am very interested in both theoretical and applied topics in learning for decision making (such as clinical trial, personalized and precision treatment, etc).
I am very foruntate to have been working with such a group of very talented researchers and students. On a daily basis, I am constantly amazed by how much you have been inspiring me.
With your great efforts, we have achieved a lot. Thank you and best wishes for your future journey!
Despite your short stay, you have broadened significantly the horizon that I can see. Thank you and I hope seeing you soon again!