Partially Observed Exchangeable Modeling

Yang Li, Junier Oliva
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:6460-6470, 2021.

Abstract

Modeling dependencies among features is fundamental for many machine learning tasks. Although there are often multiple related instances that may be leveraged to inform conditional dependencies, typical approaches only model conditional dependencies over individual instances. In this work, we propose a novel framework, partially observed exchangeable modeling (POEx) that takes in a set of related partially observed instances and infers the conditional distribution for the unobserved dimensions over multiple elements. Our approach jointly models the intra-instance (among features in a point) and inter-instance (among multiple points in a set) dependencies in data. POEx is a general framework that encompasses many existing tasks such as point cloud expansion and few-shot generation, as well as new tasks like few-shot imputation. Despite its generality, extensive empirical evaluations show that our model achieves state-of-the-art performance across a range of applications.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-li21q, title = {Partially Observed Exchangeable Modeling}, author = {Li, Yang and Oliva, Junier}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {6460--6470}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/li21q/li21q.pdf}, url = {https://proceedings.mlr.press/v139/li21q.html}, abstract = {Modeling dependencies among features is fundamental for many machine learning tasks. Although there are often multiple related instances that may be leveraged to inform conditional dependencies, typical approaches only model conditional dependencies over individual instances. In this work, we propose a novel framework, partially observed exchangeable modeling (POEx) that takes in a set of related partially observed instances and infers the conditional distribution for the unobserved dimensions over multiple elements. Our approach jointly models the intra-instance (among features in a point) and inter-instance (among multiple points in a set) dependencies in data. POEx is a general framework that encompasses many existing tasks such as point cloud expansion and few-shot generation, as well as new tasks like few-shot imputation. Despite its generality, extensive empirical evaluations show that our model achieves state-of-the-art performance across a range of applications.} }
Endnote
%0 Conference Paper %T Partially Observed Exchangeable Modeling %A Yang Li %A Junier Oliva %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-li21q %I PMLR %P 6460--6470 %U https://proceedings.mlr.press/v139/li21q.html %V 139 %X Modeling dependencies among features is fundamental for many machine learning tasks. Although there are often multiple related instances that may be leveraged to inform conditional dependencies, typical approaches only model conditional dependencies over individual instances. In this work, we propose a novel framework, partially observed exchangeable modeling (POEx) that takes in a set of related partially observed instances and infers the conditional distribution for the unobserved dimensions over multiple elements. Our approach jointly models the intra-instance (among features in a point) and inter-instance (among multiple points in a set) dependencies in data. POEx is a general framework that encompasses many existing tasks such as point cloud expansion and few-shot generation, as well as new tasks like few-shot imputation. Despite its generality, extensive empirical evaluations show that our model achieves state-of-the-art performance across a range of applications.
APA
Li, Y. & Oliva, J.. (2021). Partially Observed Exchangeable Modeling. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:6460-6470 Available from https://proceedings.mlr.press/v139/li21q.html.

Related Material