Metropolis-hastings improved particle smoother and marginalized models. Nordh, J. In 2015 23rd European Signal Processing Conference (EUSIPCO), pages 973-977, Aug, 2015. Paper doi abstract bibtex This paper combines the Metropolis-Hastings Improved Particle Smoother (MHIPS) with marginalized models. It demonstrates the effectiveness of the combination by looking at two examples; a degenerate model of a double integrator and a fifth order mixed linear/nonlinear Gaussian (MLNLG) model. For the MLNLG model two different methods are compared with the non-marginalized case; the first marginalizes the linear states only during the filtering, the second marginalizes during both the foward filtering and backward smoothing pass. The results demonstrate that marginalization not only improves the overall performance, but also increases the rate of improvement for each iteration of the MHIPS algorithm. It thus reduces the required number of iterations to beat the performance of a Forward-Filter Backward Simulator approach for the same model.
@InProceedings{7362528,
author = {J. Nordh},
booktitle = {2015 23rd European Signal Processing Conference (EUSIPCO)},
title = {Metropolis-hastings improved particle smoother and marginalized models},
year = {2015},
pages = {973-977},
abstract = {This paper combines the Metropolis-Hastings Improved Particle Smoother (MHIPS) with marginalized models. It demonstrates the effectiveness of the combination by looking at two examples; a degenerate model of a double integrator and a fifth order mixed linear/nonlinear Gaussian (MLNLG) model. For the MLNLG model two different methods are compared with the non-marginalized case; the first marginalizes the linear states only during the filtering, the second marginalizes during both the foward filtering and backward smoothing pass. The results demonstrate that marginalization not only improves the overall performance, but also increases the rate of improvement for each iteration of the MHIPS algorithm. It thus reduces the required number of iterations to beat the performance of a Forward-Filter Backward Simulator approach for the same model.},
keywords = {Gaussian processes;smoothing methods;marginalized models;metropolis-hastings improved particle smoother;MHIPS;degenerate model;double integrator;fifth order mixed linear/nonlinear Gaussian model;MLNLG model;foward filtering;backward smoothing pass;forward-filter backward simulator approach;Smoothing methods;Trajectory;Kalman filters;Proposals;Computational modeling;Europe;Metropolis-Hasting Improved Particle Smoother;Rao-Blackwellized smoothing;Particle Smoothing;Particle Filter},
doi = {10.1109/EUSIPCO.2015.7362528},
issn = {2076-1465},
month = {Aug},
url = {https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104589.pdf},
}
Downloads: 0
{"_id":"b4KdBqaYbgQmbwp9K","bibbaseid":"nordh-metropolishastingsimprovedparticlesmootherandmarginalizedmodels-2015","authorIDs":[],"author_short":["Nordh, J."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["J."],"propositions":[],"lastnames":["Nordh"],"suffixes":[]}],"booktitle":"2015 23rd European Signal Processing Conference (EUSIPCO)","title":"Metropolis-hastings improved particle smoother and marginalized models","year":"2015","pages":"973-977","abstract":"This paper combines the Metropolis-Hastings Improved Particle Smoother (MHIPS) with marginalized models. It demonstrates the effectiveness of the combination by looking at two examples; a degenerate model of a double integrator and a fifth order mixed linear/nonlinear Gaussian (MLNLG) model. For the MLNLG model two different methods are compared with the non-marginalized case; the first marginalizes the linear states only during the filtering, the second marginalizes during both the foward filtering and backward smoothing pass. The results demonstrate that marginalization not only improves the overall performance, but also increases the rate of improvement for each iteration of the MHIPS algorithm. It thus reduces the required number of iterations to beat the performance of a Forward-Filter Backward Simulator approach for the same model.","keywords":"Gaussian processes;smoothing methods;marginalized models;metropolis-hastings improved particle smoother;MHIPS;degenerate model;double integrator;fifth order mixed linear/nonlinear Gaussian model;MLNLG model;foward filtering;backward smoothing pass;forward-filter backward simulator approach;Smoothing methods;Trajectory;Kalman filters;Proposals;Computational modeling;Europe;Metropolis-Hasting Improved Particle Smoother;Rao-Blackwellized smoothing;Particle Smoothing;Particle Filter","doi":"10.1109/EUSIPCO.2015.7362528","issn":"2076-1465","month":"Aug","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104589.pdf","bibtex":"@InProceedings{7362528,\n author = {J. Nordh},\n booktitle = {2015 23rd European Signal Processing Conference (EUSIPCO)},\n title = {Metropolis-hastings improved particle smoother and marginalized models},\n year = {2015},\n pages = {973-977},\n abstract = {This paper combines the Metropolis-Hastings Improved Particle Smoother (MHIPS) with marginalized models. It demonstrates the effectiveness of the combination by looking at two examples; a degenerate model of a double integrator and a fifth order mixed linear/nonlinear Gaussian (MLNLG) model. For the MLNLG model two different methods are compared with the non-marginalized case; the first marginalizes the linear states only during the filtering, the second marginalizes during both the foward filtering and backward smoothing pass. The results demonstrate that marginalization not only improves the overall performance, but also increases the rate of improvement for each iteration of the MHIPS algorithm. It thus reduces the required number of iterations to beat the performance of a Forward-Filter Backward Simulator approach for the same model.},\n keywords = {Gaussian processes;smoothing methods;marginalized models;metropolis-hastings improved particle smoother;MHIPS;degenerate model;double integrator;fifth order mixed linear/nonlinear Gaussian model;MLNLG model;foward filtering;backward smoothing pass;forward-filter backward simulator approach;Smoothing methods;Trajectory;Kalman filters;Proposals;Computational modeling;Europe;Metropolis-Hasting Improved Particle Smoother;Rao-Blackwellized smoothing;Particle Smoothing;Particle Filter},\n doi = {10.1109/EUSIPCO.2015.7362528},\n issn = {2076-1465},\n month = {Aug},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104589.pdf},\n}\n\n","author_short":["Nordh, J."],"key":"7362528","id":"7362528","bibbaseid":"nordh-metropolishastingsimprovedparticlesmootherandmarginalizedmodels-2015","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104589.pdf"},"keyword":["Gaussian processes;smoothing methods;marginalized models;metropolis-hastings improved particle smoother;MHIPS;degenerate model;double integrator;fifth order mixed linear/nonlinear Gaussian model;MLNLG model;foward filtering;backward smoothing pass;forward-filter backward simulator approach;Smoothing methods;Trajectory;Kalman filters;Proposals;Computational modeling;Europe;Metropolis-Hasting Improved Particle Smoother;Rao-Blackwellized smoothing;Particle Smoothing;Particle Filter"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2015url.bib","creationDate":"2021-02-13T17:31:52.378Z","downloads":0,"keywords":["gaussian processes;smoothing methods;marginalized models;metropolis-hastings improved particle smoother;mhips;degenerate model;double integrator;fifth order mixed linear/nonlinear gaussian model;mlnlg model;foward filtering;backward smoothing pass;forward-filter backward simulator approach;smoothing methods;trajectory;kalman filters;proposals;computational modeling;europe;metropolis-hasting improved particle smoother;rao-blackwellized smoothing;particle smoothing;particle filter"],"search_terms":["metropolis","hastings","improved","particle","smoother","marginalized","models","nordh"],"title":"Metropolis-hastings improved particle smoother and marginalized models","year":2015,"dataSources":["eov4vbT6mnAiTpKji","knrZsDjSNHWtA9WNT"]}