Abstract: The presentation will cover our recent developments in machine learning’s application to lung cancer detection and diabetes diagnosis. Lung cancer is the leading cause of cancer deaths world-wide. Early detection of cancer is critical for therapeutic effectiveness and survival improvement. DNA methylation is known to provide potential biomarkers for assessment of cancer risk. Early diagnosis is the key to improving survival rates for lung cancer. The difficulty of detecting lung cancer is that there is subtle difference of DNA methylation level between healthy people and people with tumor. In order to increase the diagnostic performance, it is necessary to develop auxiliary diagnosis system for lung cancer using machine learning methods. A modified support vector machine (SVM) classifier is implemented for classifying lung cancer stage group. Diabetic retinopathy (DR) is an eye disease that can cause vision loss or even blindness associated with long-standing diabetes. While blindness caused by DR is mostly preventable with early detection and treatment, it is difficult to diagnose, and many patients do not get screened due to the disease’s slow progression, the lack of access to eye specialists, or simple unawareness. Automated DR screening methods with high accuracy have the strong potential to assist doctors in evaluating more patients and quickly routing those who need help to a specialist. Our current progress of constructing the deep neural networks based automated DR screening system will be presented in this talk.

Bio: Cheng-Ju Wu completed his Ph.D. program in the Department of Mechanical Engineering at University of California, Berkeley advised by Professor Roberto Horowitz. He is currently a Postdoctoral Scholar in the Department of Industrial Engineering and Operations Research at UC Berkeley mentored by Professor Xin Guo. His current research focuses on machine learning methods and its application to disease detection and diagnosis.