Predictive hypothesis identification
Description
While statistics focusses on hypothesis testing and on estimating (properties of) the true sampling distribution, in machine learning the performance of learning algorithms on future data is the primary issue. In this paper we bridge the gap with a general principle (PHI) that identifies hypotheses with best predictive performance. This includes predictive point and interval estimation, simple and composite hypothesis testing, (mixture) model selection, and others as special cases. For...[Show more]
dc.contributor.author | Hutter, Marcus | |
---|---|---|
dc.date.accessioned | 2015-08-26T06:28:29Z | |
dc.date.available | 2015-08-26T06:28:29Z | |
dc.identifier.uri | http://hdl.handle.net/1885/14970 | |
dc.description.abstract | While statistics focusses on hypothesis testing and on estimating (properties of) the true sampling distribution, in machine learning the performance of learning algorithms on future data is the primary issue. In this paper we bridge the gap with a general principle (PHI) that identifies hypotheses with best predictive performance. This includes predictive point and interval estimation, simple and composite hypothesis testing, (mixture) model selection, and others as special cases. For concrete instantiations we will recover well-known methods, variations thereof, and new ones. PHI nicely justifies, reconciles, and blends (a reparametrization invariant variation of) MAP, ML, MDL, and moment estimation. One particular feature of PHI is that it can genuinely deal with nested hypotheses. | |
dc.relation.ispartof | Ninth Valencia International Meeting on Bayesian Statistics and 2010 ISBA World Meeting | |
dc.rights | © The Author(s) | |
dc.subject | parameter estimation | |
dc.subject | hypothesis testing | |
dc.subject | model selection | |
dc.subject | predictive inference | |
dc.subject | composite hypotheses | |
dc.subject | MAP versus ML | |
dc.subject | moment fitting | |
dc.subject | Bayesian statistics | |
dc.title | Predictive hypothesis identification | |
dc.type | Conference paper | |
dc.date.issued | 2008-09 | |
local.type.status | Submitted Version | |
local.contributor.affiliation | Hutter, M., Research School of Computer Science, The Australian National University | |
local.bibliographicCitation.startpage | 1 | |
local.bibliographicCitation.lastpage | 16 | |
Collections | ANU Research Publications |
Download
File | Description | Size | Format | Image |
---|---|---|---|---|
Hutter Predictive Hypothesis Identification 2008.pdf | 254.51 kB | Adobe PDF |
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.
Updated: 17 November 2022/ Responsible Officer: University Librarian/ Page Contact: Library Systems & Web Coordinator