DocumentCode :
1450745
Title :
Bayesian function learning using MCMC methods
Author :
Magni, Paolo ; Bellazzi, Riccardo ; De Nicolao, Giuseppe
Author_Institution :
Dipartimento di Inf. e Sistemistica, Pavia Univ., Italy
Volume :
20
Issue :
12
fYear :
1998
fDate :
12/1/1998 12:00:00 AM
Firstpage :
1319
Lastpage :
1331
Abstract :
The paper deals with the problem of reconstructing a continuous 1D function from discrete noisy samples. The measurements may also be indirect in the sense that the samples may be the output of a linear operator applied to the function. Bayesian estimation provides a unified treatment of this class of problems. We show that a rigorous Bayesian solution can be efficiently implemented by resorting to a Markov chain Monte Carlo (MCMC) simulation scheme. In particular, we discuss how the structure of the problem can be exploited in order to improve the computational and convergence performances. The effectiveness of the proposed scheme is demonstrated on two classical benchmark problems as well as on the analysis of IVGTT (IntraVenous glucose tolerance test) data, a complex identification-deconvolution problem concerning the estimation of the insulin secretion rate following the administration of an intravenous glucose injection
Keywords :
Bayes methods; Markov processes; Monte Carlo methods; deconvolution; estimation theory; functional analysis; identification; inverse problems; learning (artificial intelligence); signal processing; 1D function reconstruction; Bayesian estimation; Bayesian function learning; Markov chain; Monte Carlo simulation; deconvolution; discrete noisy samples; dynamic systems; identification; insulin secretion rate; intravenous glucose injection; inverse problem; smoothing; Bayesian methods; Computational modeling; Deconvolution; Insulin; Inverse problems; Monte Carlo methods; Parameter estimation; Probability distribution; Sugar; Tellurium;
fLanguage :
English
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
Publisher :
ieee
ISSN :
0162-8828
Type :
jour
DOI :
10.1109/34.735805
Filename :
735805
Link To Document :
بازگشت