Statistics > Machine Learning
[Submitted on 11 May 2017 (this version), latest version 14 Jan 2021 (v4)]
Title:Bayesian Distribution Regression
View PDFAbstract: Distribution regression has recently attracted much interest as a generic solution to the problem of supervised learning where labels are available at the group level, rather than at the individual level. Current approaches, however, do not propagate the uncertainty in observations due to sampling variability in the groups. This effectively assumes that small and large groups are estimated equally well, and should have equal weight in the final regression. We construct a Bayesian distribution regression formalism that accounts for this uncertainty, improving the robustness and performance of the model when group sizes vary. We frame the model in a neural network style, allowing for simple MAP inference using backpropagation to learn the parameters, as well as MCMC-based inference which can fully propagate uncertainty. We demonstrate our approach on illustrative toy datasets, as well as on an astrostatistics problem in which velocity distributions are used to predict galaxy cluster masses, quantifying the distribution of dark matter in the universe.
Submission history
From: Ho Chung Leon Law [view email][v1] Thu, 11 May 2017 17:20:07 UTC (6,510 KB)
[v2] Mon, 16 Oct 2017 10:18:41 UTC (457 KB)
[v3] Thu, 22 Feb 2018 14:08:11 UTC (410 KB)
[v4] Thu, 14 Jan 2021 19:08:00 UTC (410 KB)
Current browse context:
stat.ML
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.