The Big Data for Bio Intelligence Laboratory (BDBI) aims to become a leading laboratory in the research of big data for biological intelligence and to bridge the knowledge gap between academics and practitioners. Research areas of the Laboratory include new big data solutions such as “deep learning solution”, which generates rich features to describe a machine learning problem in order to let computers make decisions, and “transductive transfer learning” – allowing computer models to be easily adapted for use in many different application domains. It will also focus on genetic farming with objectives of making the process more automatic and user-friendly and scaling it to cater to very large data sources.
1. Simple illustration of different forms of Machine Learning: supervised learning, deep learning, transfer learning, multi-task learning, reinforcement learning. Click here to see more about BDBI background.
2. Brief introduction of current research under BDBI.
From the personalized task-oriented dialogue system that is enabled from cross domain sentiment classification through reinforcement transfer learning, to more complex and powerful distant domain transfer learning. Click here to see more about current research.
3. The way forward in the future.