Mathematics > Statistics Theory
[Submitted on 7 Feb 2014 (v1), revised 21 Apr 2014 (this version, v2), latest version 26 Jan 2015 (v6)]
Title:Consistent, Two-Stage Sampled Distribution Regression via Mean Embedding
View PDFAbstract:We study the distribution regression problem: regressing to a real-valued response from a probability distribution. Due to the inherent two-stage sampled difficulty of this important machine learning problem---in practise we only have samples from sampled distributions---very little is known about its theoretical properties. In this paper, we propose an algorithmically simple approach to tackle the distribution regression problem: embed the distributions to a reproducing kernel Hilbert space, and learn a ridge regressor from the embeddings to the outputs. Our main contribution is to prove that this technique is consistent in the two-stage sampled setting under fairly mild conditions (for probability distributions on locally compact Polish spaces on which kernels have been defined). The method gives state-of-the-art results on (i) supervised entropy learning and (ii) the prediction problem of aerosol optical depth based on satellite images.
Submission history
From: Zoltan Szabo [view email][v1] Fri, 7 Feb 2014 20:37:59 UTC (58 KB)
[v2] Mon, 21 Apr 2014 11:35:58 UTC (62 KB)
[v3] Sun, 4 May 2014 19:29:36 UTC (36 KB)
[v4] Sat, 7 Jun 2014 17:42:06 UTC (51 KB)
[v5] Sat, 25 Oct 2014 21:03:01 UTC (57 KB)
[v6] Mon, 26 Jan 2015 22:20:59 UTC (57 KB)
Current browse context:
math.ST
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
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.