Cultural advice

The Australian National University acknowledges, celebrates and pays our respects to the Ngunnawal and Ngambri people of the Canberra region and to all First Nations Australians on whose traditional lands we meet and work, and whose cultures are among the oldest continuing cultures in human history.

Aboriginal and Torres Strait Islander peoples are advised that ANU Library collections may include images, names, voices, and other representations of deceased persons.

Material in the collection may contain terms, language or views that reflect the period in which the item was created and may be considered inappropriate today.

Discrete MDL predicts in total variation

Simple item page
dc.contributor.authorHutter, Marcus
dc.coverage.spatialVancouver Canada
dc.date.accessioned2015-12-10T22:38:46Z
dc.date.createdDecember 7-12 2009
dc.date.issued2009
dc.date.updated2016-02-24T11:44:32Z
dc.description.abstractThe Minimum Description Length (MDL) principle selects the model that has the shortest code for data plus model. We show that for a countable class of models, MDL predictions are close to the true distribution in a strong sense. The result is completely general. No independence, ergodicity, stationarity, identifiability, or other assumption on the model class need to be made. More formally, we show that for any countable class of models, the distributions selected by MDL (or MAP) asymptotically predict (merge with) the true measure in the class in total variation distance. Implications for non-i.i.d. domains like time-series forecasting, discriminative learning, and reinforcement learning are discussed.
dc.identifier.urihttp://hdl.handle.net/1885/56874
dc.publisherMIT Press
dc.relation.ispartofseriesConference on Advances in Neural Information Processing Systems (NIPS 2009)
dc.rightsCopyright Information: © The Author(s)
dc.sourceProceedings of The 23rd Annual Conference on Neural Information Processing Systems (NIPS 23)
dc.source.urihttp://books.nips.cc/nips22.html
dc.subjectKeywords: Discriminative learning; Ergodicity; Identifiability; Minimum description length principle; Stationarity; Time series forecasting; Total variation; Reinforcement learning
dc.titleDiscrete MDL predicts in total variation
dc.typeConference paper
local.bibliographicCitation.lastpage825
local.bibliographicCitation.startpage817
local.contributor.affiliationHutter, Marcus, College of Engineering and Computer Science, ANU
local.contributor.authoruidHutter, Marcus, u4350841
local.description.embargo2037-12-31
local.description.notesImported from ARIES
local.description.refereedYes
local.identifier.absfor080401 - Coding and Information Theory
local.identifier.absfor080101 - Adaptive Agents and Intelligent Robotics
local.identifier.absfor010405 - Statistical Theory
local.identifier.ariespublicationu8803936xPUB378
local.identifier.scopusID2-s2.0-84858716044
local.type.statusPublished Version

Downloads

Original bundle

Now showing 1 - 2 of 2
Thumbnail Image
Name:
01_Hutter_Discrete_MDL_predicts_in_total_2009.pdf
Size:
286.16 KB
Format:
Adobe Portable Document Format
Download
Thumbnail Image
Name:
02_Hutter_Discrete_MDL_predicts_in_total_2009.pdf
Size:
16.13 KB
Format:
Adobe Portable Document Format
Download

Collections

ANU Research Publications

DSpace software copyright © 2002-2026 LYRASIS