The advancement in deriving powerful data-driven representation has been key to the recent success across machine intelligence tasks. This representation embeds information in a high dimensional feature vector capturing the essence of complex and non-linear structure of our data space. In this talk, we will demonstrate the effectiveness of representational learning in the context of AI-assisted diagnosis application. A phenotype representation is learned using a deep generative model. Then, jointly with the National Taiwan University Hospital, we developed an AI-assisted interpretation algorithm of leukemia on over 10,000 unique patient’s flow cytometry data – achieving a remarkable accuracy >0.9 AUC.
Chi-Chun Lee (Jeremy) is an Assistant Professor at the Electrical Engineering Department of the National Tsing Hua University (NTHU), Taiwan. He received his Ph.D. degree in Electrical Engineering from the University of Southern California, USA in 2012. He was a data scientist at id:a lab at ID Analytics in 2013. His research interests are in the algorithmic development for human-centered behavioral signal processing (BSP) and affective computing. He has been involved in multiple granted interdisciplinary research projects, including aspects on education, psychology, neuroscience, and clinical health applications, with a focus on deriving decision analytics using signal processing and machine learning.