I'm now a fourth-year PhD candidate at the National University of Singapore supervised by Prof. Dong Jin Song. I received my Bachelor from Fudan University in 2016.
My research interests lie in robust and trustworthy machine learning techniques, emphasizing understanding, debugging, and ensuring the behavior of machine learning models; in particular, my work focuses on developing a visual interactive system that vividly depicts the internal evolution of deep neural networks, allowing for the inspection of ML pipelines and the identification of potential issues.
I'm open to discussion or collaboration. Feel free to drop me an email if you're interested in my research.
Understanding how the predictions of deep learning models are formed during the training process is crucial to improve model performance and fix model defects, especially when we need to investigate nontrivial training strategies such as active learning, and track the root cause of unexpected training results such as performance degeneration. In this work, we propose a time-travelling visual solution DeepVisualInsight (DVI), aiming to manifest the spatio-temporal causality while training a deep learning image classifier. The spatio-temporal causality demonstrates how the gradient-descent algorithm and various training data sampling techniques can influence and reshape the layout of learnt input representation and the classification boundaries in consecutive epochs. Such causality allows us to observe and analyze the whole learning process in the visible low dimensional space. Technically, we propose four spatial and temporal properties and design our visualization solution to satisfy them. These properties preserve the most important information when (inverse-)projecting input samples between the visible low-dimensional and the invisible high-dimensional space, for causal analyses. Our extensive experiments show that, comparing to baseline approaches, we achieve the best visualization performance regarding the spatial/temporal properties and visualization efficiency. Moreover, our case study shows that our visual solution can well reflect the characteristics of various training scenarios, showing good potential of DVI as a debugging tool for analyzing deep learning training processes.
Computing 2,
15 Computing Drive,
National University of Singapore,
Singapore, 117418
Email: xianglin[at]u[dot]nus[dot]edu
© 2022 YANG Xianglin | Powered by Skeleton and Jiarui Gan's template | Updated 2022-05-16