Please use this identifier to cite or link to this item: http://localhost/handle/Hannan/602743
Title: Latency probability estimation of non-linear systems with one-step randomly delayed measurements
Authors: Yulong Huang;Yonggang Zhang;Ning Li
subject: unknown constant latency probability estimation algorithm|Newton approach|M-step|expectation maximisation approach|maximisation step|univariate nonstationary growth model|maximum likelihood criterion|complete data log-likelihood function|ML criterion|one-step randomly delayed measurements|E-step|nonlinear systems|expectation step
Year: 2016
Publisher: IEEE
Abstract: In this study, the authors focus on estimating the unknown constant latency probability of non-linear systems with one-step randomly delayed measurements using maximum likelihood (ML) criterion. A new latency probability estimation algorithm is proposed based on an expectation maximisation approach to obtain an approximate ML estimation of latency probability. The proposed algorithm consists of expectation step (E-step) and the maximisation step (M-step). In the E-step, the expectation of the complete data log-likelihood function is approximately computed based on the currently estimated latency probability, and in the M-step, the approximate expectation is maximised using the Newton approach. The efficacy of the proposed algorithm is illustrated in a numerical example concerning univariate non-stationary growth model.
URI: http://localhost/handle/Hannan/147822
http://localhost/handle/Hannan/602743
ISSN: 1751-8644
1751-8652
volume: 10
issue: 7
Appears in Collections:2016

Files in This Item:
File Description SizeFormat 
7449097.pdf901.3 kBAdobe PDFThumbnail
Preview File
Title: Latency probability estimation of non-linear systems with one-step randomly delayed measurements
Authors: Yulong Huang;Yonggang Zhang;Ning Li
subject: unknown constant latency probability estimation algorithm|Newton approach|M-step|expectation maximisation approach|maximisation step|univariate nonstationary growth model|maximum likelihood criterion|complete data log-likelihood function|ML criterion|one-step randomly delayed measurements|E-step|nonlinear systems|expectation step
Year: 2016
Publisher: IEEE
Abstract: In this study, the authors focus on estimating the unknown constant latency probability of non-linear systems with one-step randomly delayed measurements using maximum likelihood (ML) criterion. A new latency probability estimation algorithm is proposed based on an expectation maximisation approach to obtain an approximate ML estimation of latency probability. The proposed algorithm consists of expectation step (E-step) and the maximisation step (M-step). In the E-step, the expectation of the complete data log-likelihood function is approximately computed based on the currently estimated latency probability, and in the M-step, the approximate expectation is maximised using the Newton approach. The efficacy of the proposed algorithm is illustrated in a numerical example concerning univariate non-stationary growth model.
URI: http://localhost/handle/Hannan/147822
http://localhost/handle/Hannan/602743
ISSN: 1751-8644
1751-8652
volume: 10
issue: 7
Appears in Collections:2016

Files in This Item:
File Description SizeFormat 
7449097.pdf901.3 kBAdobe PDFThumbnail
Preview File
Title: Latency probability estimation of non-linear systems with one-step randomly delayed measurements
Authors: Yulong Huang;Yonggang Zhang;Ning Li
subject: unknown constant latency probability estimation algorithm|Newton approach|M-step|expectation maximisation approach|maximisation step|univariate nonstationary growth model|maximum likelihood criterion|complete data log-likelihood function|ML criterion|one-step randomly delayed measurements|E-step|nonlinear systems|expectation step
Year: 2016
Publisher: IEEE
Abstract: In this study, the authors focus on estimating the unknown constant latency probability of non-linear systems with one-step randomly delayed measurements using maximum likelihood (ML) criterion. A new latency probability estimation algorithm is proposed based on an expectation maximisation approach to obtain an approximate ML estimation of latency probability. The proposed algorithm consists of expectation step (E-step) and the maximisation step (M-step). In the E-step, the expectation of the complete data log-likelihood function is approximately computed based on the currently estimated latency probability, and in the M-step, the approximate expectation is maximised using the Newton approach. The efficacy of the proposed algorithm is illustrated in a numerical example concerning univariate non-stationary growth model.
URI: http://localhost/handle/Hannan/147822
http://localhost/handle/Hannan/602743
ISSN: 1751-8644
1751-8652
volume: 10
issue: 7
Appears in Collections:2016

Files in This Item:
File Description SizeFormat 
7449097.pdf901.3 kBAdobe PDFThumbnail
Preview File