Fast identification algorithms for Gaussian process model

Lists

Download

Full text not archived in this repository.

Tools

Hong, X. ORCID: https://orcid.org/0000-0002-6832-2298, Gao, J., Jiang, X. and Harris, C. J. (2014) Fast identification algorithms for Gaussian process model. Neurocomputing, 133. pp. 25-31. ISSN 0925-2312 doi: 10.1016/j.neucom.2013.11.035

Abstract/Summary

A class identification algorithms is introduced for Gaussian process(GP)models.The fundamental approach is to propose a new kernel function which leads to a covariance matrix with low rank,a property that is consequently exploited for computational efficiency for both model parameter estimation and model predictions.The objective of either maximizing the marginal likelihood or the Kullback–Leibler (K–L) divergence between the estimated output probability density function(pdf)and the true pdf has been used as respective cost functions.For each cost function,an efficient coordinate descent algorithm is proposed to estimate the kernel parameters using a one dimensional derivative free search, and noise variance using a fast gradient descent algorithm. Numerical examples are included to demonstrate the effectiveness of the new identification approaches.

Altmetric Badge

Item Type	Article
URI	https://reading-clone.eprints-hosting.org/id/eprint/36487
Item Type	Article
Refereed	Yes
Divisions	Science > School of Mathematical, Physical and Computational Sciences > Department of Computer Science
Uncontrolled Keywords	Gaussian process Optimization Kullback–Leibler divergence
Publisher	Elsevier
Download/View statistics	View download statistics for this item

Related URLs

Deposit Details

Date Deposited:	24 Apr 2014 10:04	Date item deposited into CentAUR
Last Modified:	09 Jun 2024 01:50	Date item last modified

References

[1] C.E.Rasmussen, “Gaussian processesinmachinelearning,” in: Advanced Lectures onMachineLearning.LectureNotesinComputerScience,vol.3176, 2004,pp.63–71. [2] C.C.E. Rasmussen,C.K.I.Williams,GaussianProcessesforMachineLearning, MIT Press,USA,2006. [3] N. Lawrence,Probabilisticnon-linearprincipalcomponentanalysiswith Gaussian processlatentvariablemodels,J.Mach.Learn.Res.6(2005) 1783–1816. [4] X. Jiang,J.Gao,T.Wang,L.Zheng,SupervisedlatentlinearGaussianprocess latent variablemodelfordimensionalityreduction,IEEETrans.Syst.Man Cybern. PartB:Cybern.42(6)(2012)1620–1632. [5] R.D. Turner,M.F.Huber,U.D.Hanebeck,C.E.Rasmussen,Robust filtering and smoothing withGaussianprocesses,IEEETrans.Autom.Control57(7)(2012) 1865–1871. [6] B. Schölkopf,A.Smola,LearningwithKernels,TheMITPress,Cambridge,MA, 2002. [7] S. Kullback,R.A.Leibler,Oninformationandsufficiency, Ann.Math.Stat.22(1) (1951)79–86. [8] K.P.Burnham,D.R.Anderson,ModelSelectionandMultimodelInference: A PracticalInformation-TheoreticApproach,SpringerScience,NewYork,2002. [9] X.Hong,J.Gao,X.Jiang,C.Harris,EstimationofGaussianprocessregression model usingprobabilitydistancemeasures,submittedtojournal"Systems Science &ControlEngineering". [10] J. QuiñoneroCandela,C.E.Rasmussen,Aunifyingviewofsparseapproximate gaussian processregression,J.Mach.Learn.Res.6(2005)1939–1959. [11]C.Williams,M.Seeger,Usingthenyströnmethodtospeedupkernel machines, in:AdvancesinNeuralInformationProcessingSystems13,MIT Press, USA,2001,pp.682–688. [12] S.A.Billings,W.Voon,Aprediction-errorandstepwise-regressionestimation algorithm fornonlinearsystems,Int.J.Control44(1986)803–822. [13] A.W.vanderVaart,AsymptoticStatistics,CambridgePress,UK,2000. [14] S. Chen,X.Hong,C.J.Harris,P.M.Sharkey,Sparsemodellingusingorthogonal forwardregressionwithPRESSstatisticandregularization,IEEETrans.Syst. Man Cybern.PartB:Cybern.34(2)(2004)898–911. [15] S. Chen,X.Hong,C.J.Harris,Constructionoftunableradialbasisfunction networksusingorthogonalforwardselection,IEEETrans.Syst.ManCybern. Part B:Cybern.39(2)(2009)457–466.

University Staff: Request a correction | Centaur Editors: Update this record

Search Google Scholar