Partial observations of continuous time-series dynamics at arbitrary time stamps exist in many disciplines. Fitting this type of data using statistical models with continuous dynamics is not only promising at an intuitive level but also has practical benefits, including the ability to generate continuous trajectories and to perform inference on previously unseen time stamps. Despite exciting progress in this area, the existing models still face challenges in terms of their representational power and the quality of their variational approximations. We tackle these challenges with continuous latent process flows (CLPF), a principled architecture decoding continuous latent processes into continuous observable processes using a time-dependent normalizing flow driven by a stochastic differential equation. To optimize our model using maximum likelihood, we propose a novel piecewise construction of a variational posterior process and derive the corresponding variational lower bound using trajectory re-weighting. Our ablation studies demonstrate the effectiveness of our contributions in various inference tasks on irregular time grids. Comparisons to state-of-the-art baselines show our model's favourable performance on both synthetic and real-world time-series data.
Ruizhi Deng (Simon Fraser University)
Ruizhi Deng is a Master of Science student in computing science at Simon Fraser University. He works in VML lab and he is advised by [Dr. Greg Mori](http://www.cs.sfu.ca/~mori/). He's interested in studying fundamental problems in machine learning, especially deep learning. His recent research focus is adversarial machine learning. He also has past and on-going experience in designing network architectures and developing generative models in computer vision. Before coming to Simon Fraser University, he obtained his Bachelor of Science degree from the University of Michigan - Ann Arbor. His research advisor was [Dr. Honglak Lee](http://web.eecs.umich.edu/~honglak/).
Marcus Brubaker (York University)
Greg Mori (Borealis AI)
Andreas M Lehrmann (Disney Research)
More from the Same Authors
2020 Poster: Modeling Continuous Stochastic Processes with Dynamic Normalizing Flows »
Ruizhi Deng · Bo Chang · Marcus Brubaker · Greg Mori · Andreas Lehrmann
2020 Poster: Wavelet Flow: Fast Training of High Resolution Normalizing Flows »
Jason Yu · Konstantinos Derpanis · Marcus Brubaker