Timezone: »

 
Invertible Neural Networks for Graph Prediction
Chen Xu · Xiuyuan Cheng · Yao Xie
Event URL: https://openreview.net/forum?id=ErTlGsDVDbi »
In this work, we address conditional generation using deep invertible neural networks. This is a type of problem where one aims to infer the most probable inputs $X$ given outcomes $Y$. We call our method \textit{invertible graph neural network} (iGNN) due to the primary focus on generating node features on graph data. A notable feature of our proposed methods is that during network training, we revise the typically-used loss objective in normalizing flow and consider Wasserstein-2 regularization to facilitate the training process. Algorithmic-wise, we adopt an end-to-end training approach since our objective is to address prediction and generation in the forward and backward processes at once through a single model. Theoretically, we study the expressiveness of iGNN in learning the mapping through utilizing the Fokker-Planck equation of an Ornstein-Uhlenbeck process. Experimentally, we verify the performance of iGNN on both simulated and real-data datasets. We demonstrate through extensive numerical experiments that iGNN shows clear improvement over competing conditional generation benchmarks on high-dimensional and/or non-convex data.

Author Information

Chen Xu (Georgia Institute of Technology)
Xiuyuan Cheng (Duke University)
Yao Xie (Georgia Institute of Technology)

Yao Xie is an Assistant Professor and Harold R. and Mary Anne Nash Early Career Professor in the H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology, which she joined in 2013. She received her Ph.D. in Electrical Engineering (minor in Mathematics) from Stanford University in 2011, M.Sc. in Electrical and Computer Engineering from the University of Florida, and B.Sc. in Electrical Engineering and Computer Science from University of Science and Technology of China (USTC) . From 2012 to 2013, she was a Research Scientist at Duke University. Her research areas include statistics, signal processing, and machine learning, in providing theoretical foundation as well as developing computationally efficient and statistically powerful algorithms for big data in various applications such as sensor networks, imaging, and crime data analysis. She received the National Science Foundation CAREER Award in 2017 and her crime data analytics project received the Smart 50 Award at the Smart Cities Connect Conferences and Expo in 2018.

More from the Same Authors