Doctoral Seminar: Qicheng Lao

Speaker: Qicheng Lao

Supervisor: Dr. T. Fevens

Supervisory Committee: Drs. M. Amer, T. D. Bui, A. Krzyzak

Title: Learning Through Image Sequences

Date: Thursday, March 14, 2019

Time: 11:40 a.m.

Place: EV 3.309

ABSTRACT

Many machine learning systems for artificial intelligence are biologically inspired, for example, the artificial neural networks (ANNs) have similar architecture as human brains, and convolutional neural networks (CNNs) are inspired by the observations from early study on animal's visual cortex system. The above two examples (ANNs and CNNs) are inspirations at the level of creating fundamental tools (e.g., neural networks) for a machine learning system. Another level of inspiration can come from the way humans learn or respond that builds on top of existing powerful learning tools, i.e., brains.

In this seminar, we will focus on another type of inspiration that also belongs to the second level. It is based on the common practice that for an efficient learning or an optimal decision, human integrate all sources of given information in multiple views and leverage the reasoning of connections among them, i.e., multi-view learning. We address a specific type of medical image recognition problems from the perspective of multi-view learning, by proposing disease progression learning to emphasize the learning of underlying connections among multiple stages of a disease, with each stage being a sequential view of the disease. Our proposed method is evaluated on a public diabetic retinopathy dataset, and achieves about 3.3% improvement in disease staging accuracy, compared to the baseline method that does not use disease progression learning.