Publication: Parallel sequential Monte Carlo for stochastic gradient-free nonconvex optimization
dc.affiliation.dpto | UC3M. Departamento de Teoría de la Señal y Comunicaciones | es |
dc.affiliation.grupoinv | UC3M. Grupo de Investigación: Tratamiento de la Señal y Aprendizaje (GTSA) | es |
dc.contributor.author | Akyildiz, Omer Deniz | |
dc.contributor.author | Crisan, Dan | |
dc.contributor.author | Míguez Arenas, Joaquín | |
dc.contributor.funder | Comunidad de Madrid | es |
dc.contributor.funder | Agencia Estatal de Investigación (España) | es |
dc.date.accessioned | 2023-07-14T07:24:35Z | |
dc.date.available | 2023-07-14T07:24:35Z | |
dc.date.issued | 2020-11 | |
dc.description.abstract | We introduce and analyze a parallel sequential Monte Carlo methodology for the numerical solution of optimization problems that involve the minimization of a cost function that consists of the sum of many individual components. The proposed scheme is a stochastic zeroth-order optimization algorithm which demands only the capability to evaluate small subsets of components of the cost function. It can be depicted as a bank of samplers that generate particle approximations of several sequences of probability measures. These measures are constructed in such a way that they have associated probability density functions whose global maxima coincide with the global minima of the original cost function. The algorithm selects the best performing sampler and uses it to approximate a global minimum of the cost function. We prove analytically that the resulting estimator converges to a global minimum of the cost function almost surely and provide explicit convergence rates in terms of the number of generated Monte Carlo samples and the dimension of the search space. We show, by way of numerical examples, that the algorithm can tackle cost functions with multiple minima or with broad "flat" regions which are hard to minimize using gradient-based techniques. | en |
dc.description.sponsorship | This work was partially supported by Agencia Estatal de Investigación of Spain (RTI2018-099655-B-I00 CLARA), and the regional government of Madrid (program CASICAM-CM S2013/ICE-2845). The work of the second author has been partially supported by a UC3M-Santander Chair of Excellence grant held at the Universidad Carlos III de Madrid. | en |
dc.format.extent | 19 | |
dc.identifier.bibliographicCitation | Akyildiz, Ö. D., Crisan, D., & Míguez, J. (2020). Parallel sequential Monte Carlo for stochastic gradient-free nonconvex optimization. Statistics and Computing, 30(6), 1645-1663. | en |
dc.identifier.doi | https://doi.org/10.1007/s11222-020-09964-4 | |
dc.identifier.issn | 0960-3174 | |
dc.identifier.publicationfirstpage | 1645 | |
dc.identifier.publicationissue | 6 | |
dc.identifier.publicationlastpage | 1663 | |
dc.identifier.publicationtitle | Statistics and Computing | en |
dc.identifier.publicationvolume | 30 | |
dc.identifier.uri | https://hdl.handle.net/10016/37838 | |
dc.identifier.uxxi | AR/0000027719 | |
dc.language.iso | eng | |
dc.publisher | Springer | en |
dc.relation.projectID | Comunidad de Madrid. S2013/ICE-2845 | es |
dc.relation.projectID | Gobierno de España. RTI2018-099655-B-I00 | es |
dc.rights | © The Author(s) 2020. | en |
dc.rights | Atribución 3.0 España | * |
dc.rights.accessRights | open access | en |
dc.rights.uri | http://creativecommons.org/licenses/by/3.0/es/ | * |
dc.subject.eciencia | Biología y Biomedicina | es |
dc.subject.eciencia | Ciencias de la Información | es |
dc.subject.eciencia | Economía | es |
dc.subject.eciencia | Electrónica | es |
dc.subject.eciencia | Estadística | es |
dc.subject.eciencia | Telecomunicaciones | es |
dc.subject.other | Sequential Monte Carlo | en |
dc.subject.other | Stochastic optimization | en |
dc.subject.other | Nonconvex optimization | en |
dc.subject.other | Gradient-free optimization | en |
dc.subject.other | Sampling | en |
dc.title | Parallel sequential Monte Carlo for stochastic gradient-free nonconvex optimization | en |
dc.type | research article | * |
dc.type.hasVersion | VoR | * |
dspace.entity.type | Publication |
Files
Original bundle
1 - 1 of 1