schliessen

Filtern

 

Bibliotheken

Model selection for sequential designs in discrete finite systems using Bernstein kernels

We view sequential design as a model selection problem to determine which new observation is expected to be the most informative, given the existing set of observations. For estimating a probability distribution on a bounded interval, we use bounds constructed from kernel density estimators along wi... Full description

Main Author: Nath, Madhurima
Other Authors: Eubank, Stephen
Format: Electronic Article Electronic Article
Language:
Subjects:
Quelle: Cornell University
ID: Arxiv ID: 1807.06661
Zum Text:
SendSend as email Add to Book BagAdd to Book Bag
Staff View
recordid: arxiv1807.06661
title: Model selection for sequential designs in discrete finite systems using Bernstein kernels
format: Article
creator:
  • Nath, Madhurima
  • Eubank, Stephen
subjects:
  • Statistics - Methodology
ispartof:
description: We view sequential design as a model selection problem to determine which new observation is expected to be the most informative, given the existing set of observations. For estimating a probability distribution on a bounded interval, we use bounds constructed from kernel density estimators along with the estimated density itself to estimate the information gain expected from each observation. We choose Bernstein polynomials for the kernel functions because they provide a complete set of basis functions for polynomials of finite degree and thus have useful convergence properties. We illustrate the method with applications to estimating network reliability polynomials, which give the probability of certain sets of configurations in finite, discrete stochastic systems.
language:
source: Cornell University
identifier: Arxiv ID: 1807.06661
fulltext: fulltext_linktorsrc
url: Link


@attributes
ID1007570731
RANK0.07
NO1
SEARCH_ENGINEprimo_central_multiple_fe
SEARCH_ENGINE_TYPEPrimo Central Search Engine
LOCALfalse
PrimoNMBib
record
control
sourcerecordid1807.06661
sourceidarxiv
recordidTN_arxiv1807.06661
sourcesystemPC
display
typearticle
titleModel selection for sequential designs in discrete finite systems using Bernstein kernels
creatorNath, Madhurima ; Eubank, Stephen
identifierArxiv ID: 1807.06661
subjectStatistics - Methodology
descriptionWe view sequential design as a model selection problem to determine which new observation is expected to be the most informative, given the existing set of observations. For estimating a probability distribution on a bounded interval, we use bounds constructed from kernel density estimators along with the estimated density itself to estimate the information gain expected from each observation. We choose Bernstein polynomials for the kernel functions because they provide a complete set of basis functions for polynomials of finite degree and thus have useful convergence properties. We illustrate the method with applications to estimating network reliability polynomials, which give the probability of certain sets of configurations in finite, discrete stochastic systems.
sourceCornell University
oafree_for_read
links
openurl$$Topenurl_article
linktorsrc$$Uhttp://arxiv.org/abs/1807.06661$$EView_record_at _arXiv
openurlfulltext$$Topenurlfull_article
search
creatorcontrib
0Nath, Madhurima
1Eubank, Stephen
titleModel selection for sequential designs in discrete finite systems using Bernstein kernels
descriptionWe view sequential design as a model selection problem to determine which new observation is expected to be the most informative, given the existing set of observations. For estimating a probability distribution on a bounded interval, we use bounds constructed from kernel density estimators along with the estimated density itself to estimate the information gain expected from each observation. We choose Bernstein polynomials for the kernel functions because they provide a complete set of basis functions for polynomials of finite degree and thus have useful convergence properties. We illustrate the method with applications to estimating network reliability polynomials, which give the probability of certain sets of configurations in finite, discrete stochastic systems.
subjectStatistics - Methodology
general
0Cornell University
1Arxiv ID: 1807.06661
sourceidarxiv
recordidarxiv1807.06661
rsrctypearticle
creationdate
020180717
12018
recordtypearticle
searchscope
0arxiv
1Arxiv
scope
0arxiv
1Arxiv
sort
titleModel selection for sequential designs in discrete finite systems using Bernstein kernels
authorNath, Madhurima ; Eubank, Stephen
creationdate20180717
facets
frbrgroupid-7662886838926867073
frbrtype6
newrecords20180724
creationdate2018
topicStatistics–Methodology
collectionarXiv
prefilterarticles
rsrctypearticles
creatorcontrib
0Nath, Madhurima
1Eubank, Stephen
delivery
delcategoryRemote Search Resource
fulltextfulltext_linktorsrc
addata
aulast
0Nath
1Eubank
aufirst
0Madhurima
1Stephen
au
0Nath, Madhurima
1Eubank, Stephen
atitleModel selection for sequential designs in discrete finite systems using Bernstein kernels
risdate20180717
genrearticle
ristypeJOUR
abstractWe view sequential design as a model selection problem to determine which new observation is expected to be the most informative, given the existing set of observations. For estimating a probability distribution on a bounded interval, we use bounds constructed from kernel density estimators along with the estimated density itself to estimate the information gain expected from each observation. We choose Bernstein polynomials for the kernel functions because they provide a complete set of basis functions for polynomials of finite degree and thus have useful convergence properties. We illustrate the method with applications to estimating network reliability polynomials, which give the probability of certain sets of configurations in finite, discrete stochastic systems.
lad21arXiv.org:1807.06661
oafree_for_read
date2018-07-17