Learning To Stop While Learning To Predict

Xinshi Chen, Hanjun Dai, Yu Li, Xin Gao, Le Song

Research output: Chapter in Book/Report/Conference proceedingConference contribution

21 Scopus citations

Abstract

There is a recent surge of interest in designing deep architectures based on the update steps in traditional algorithms, or learning neural networks to improve and replace traditional algorithms. While traditional algorithms have certain stopping criteria for outputting results at different iterations, many algorithm-inspired deep models are restricted to a "fixed-depth" for all inputs. Similar to algorithms, the optimal depth of a deep architecture may be different for different input instances, either to avoid "over-thinking", or because we want to compute less for operations converged already. In this paper, we tackle this varying depth problem using a steerable architecture, where a feed-forward deep model and a variational stopping policy are learned together to sequentially determine the optimal number of layers for each input instance. Training such architecture is very challenging. We provide a variational Bayes perspective and design a novel and effective training procedure which decomposes the task into an oracle model learning stage and an imitation stage. Experimentally, we show that the learned deep model along with the stopping policy improves the performances on a diverse set of tasks, including learning sparse recovery, few-shot meta learning, and computer vision tasks.
Original languageEnglish (US)
Title of host publicationInternational Conference on Machine Learning (ICML)
StatePublished - 2020

Bibliographical note

KAUST Repository Item: Exported on 2021-09-16
Acknowledged KAUST grant number(s): BAS/1/1624-01, FCC/1/1976-18-01, FCC/1/1976-23-01, FCC/1/1976-25-01, FCC/1/1976-26-01, REI/1/0018-01-01, REI/1/0018-01-01
Acknowledgements: We would like to thank anonymous reviewers for providing constructive feedbacks. This work is supported in part by NSF grants CDS&E-1900017 D3SC, CCF-1836936 FMitF, IIS-1841351, CAREER IIS-1350983 to L.S. and grants from King Abdullah University of Science and Technology, under award numbers BAS/1/1624-01, FCC/1/1976-18-01, FCC/1/1976-23-01, FCC/1/1976-25-01, FCC/1/1976-26-01, REI/1/0018-01-01, and URF/1/4098-01-01.

Fingerprint

Dive into the research topics of 'Learning To Stop While Learning To Predict'. Together they form a unique fingerprint.

Cite this