Generalized filtering (original) (raw)
Generalized filtering is a generic Bayesian filtering scheme for nonlinear state-space models. It is based on a variational principle of least action, formulated in generalized coordinates of motion. Note that "generalized coordinates of motion" are related to—but distinct from—generalized coordinates as used in (multibody) dynamical systems analysis. Generalized filtering furnishes posterior densities over hidden states (and parameters) generating observed data using a generalized gradient descent on variational free energy, under the Laplace assumption. Unlike classical (e.g. Kalman-Bucy or particle) filtering, generalized filtering eschews Markovian assumptions about random fluctuations. Furthermore, it operates online, assimilating data to approximate the posterior density over unknown
| Property | Value |
|---|---|
| dbo:abstract | El filtrado generalizado es un esquema de genérico para modelos de espacio de estado no lineales. Se basa en un principio variacional de mínima acción, formulado en coordenadas generalizadas. Tenga en cuenta que el concepto de "coordenadas generalizadas" como se usa aquí difiere del concepto de coordenadas generalizadas de movimiento como se usa en el análisis de sistemas dinámicos (multicuerpo). El filtrado generalizado proporciona densidades posteriores sobre estados ocultos (y parámetros) que generan datos observados utilizando un descenso de gradiente generalizado en energía libre variacional, bajo el . A diferencia del filtrado clásico (p. ej. Kalman-Bucy o de partículas), el filtrado generalizado evita las suposiciones markovianas sobre fluctuaciones aleatorias. Además, opera en línea, asimilando datos para aproximar la densidad posterior sobre cantidades desconocidas, sin la necesidad de pasos hacia atrás. Los casos especiales incluyen filtrado variacional, maximización dinámica de expectativas y codificación predictiva generalizada. (es) Generalized filtering is a generic Bayesian filtering scheme for nonlinear state-space models. It is based on a variational principle of least action, formulated in generalized coordinates of motion. Note that "generalized coordinates of motion" are related to—but distinct from—generalized coordinates as used in (multibody) dynamical systems analysis. Generalized filtering furnishes posterior densities over hidden states (and parameters) generating observed data using a generalized gradient descent on variational free energy, under the Laplace assumption. Unlike classical (e.g. Kalman-Bucy or particle) filtering, generalized filtering eschews Markovian assumptions about random fluctuations. Furthermore, it operates online, assimilating data to approximate the posterior density over unknown quantities, without the need for a backward pass. Special cases include variational filtering, dynamic expectation maximization and generalized predictive coding. (en) |
| dbo:wikiPageExternalLink | http://www.fil.ion.ucl.ac.uk/spm/software/ http://www.fil.ion.ucl.ac.uk/~karl/%23_Bayesian_and_Variational |
| dbo:wikiPageID | 39756603 (xsd:integer) |
| dbo:wikiPageLength | 18309 (xsd:nonNegativeInteger) |
| dbo:wikiPageRevisionID | 1069927937 (xsd:integer) |
| dbo:wikiPageWikiLink | dbr:Principle_of_least_action dbr:Particle_filter dbr:Dynamic_Bayesian_network dbr:Entropy_(information_theory) dbr:Generalized_coordinates dbr:Stochastic_processes dbr:Fundamental_lemma_of_calculus_of_variations dbr:Autocorrelation dbc:Nonlinear_filters dbc:Systems_theory dbr:Tuple dbr:Linear_predictive_coding dbr:Linearization dbc:Control_theory dbc:Signal_estimation dbr:Kalman_filter dbr:Taylor's_theorem dbr:Dynamic_causal_modelling dbr:Marginal_likelihood dbc:Linear_filters dbc:Markov_models dbr:Laplace's_method dbc:Bayesian_estimation dbr:Kullback–Leibler_divergence dbr:Newton's_method_in_optimization dbr:Optimal_control dbc:Stochastic_differential_equations dbr:System_identification dbr:Variational_Bayesian_methods dbr:Recursive_Bayesian_estimation |
| dbp:wikiPageUsesTemplate | dbt:Reflist |
| dct:subject | dbc:Nonlinear_filters dbc:Systems_theory dbc:Control_theory dbc:Signal_estimation dbc:Linear_filters dbc:Markov_models dbc:Bayesian_estimation dbc:Stochastic_differential_equations |
| gold:hypernym | dbr:Bayesian |
| rdf:type | yago:WikicatLinearFilters yago:WikicatNonlinearFilters yago:Artifact100021939 yago:Device103183080 yago:Filter103339643 yago:Instrumentality103575240 yago:Object100002684 yago:PhysicalEntity100001930 yago:System104377057 yago:Whole100003553 yago:WikicatSystems |
| rdfs:comment | Generalized filtering is a generic Bayesian filtering scheme for nonlinear state-space models. It is based on a variational principle of least action, formulated in generalized coordinates of motion. Note that "generalized coordinates of motion" are related to—but distinct from—generalized coordinates as used in (multibody) dynamical systems analysis. Generalized filtering furnishes posterior densities over hidden states (and parameters) generating observed data using a generalized gradient descent on variational free energy, under the Laplace assumption. Unlike classical (e.g. Kalman-Bucy or particle) filtering, generalized filtering eschews Markovian assumptions about random fluctuations. Furthermore, it operates online, assimilating data to approximate the posterior density over unknown (en) El filtrado generalizado es un esquema de genérico para modelos de espacio de estado no lineales. Se basa en un principio variacional de mínima acción, formulado en coordenadas generalizadas. Tenga en cuenta que el concepto de "coordenadas generalizadas" como se usa aquí difiere del concepto de coordenadas generalizadas de movimiento como se usa en el análisis de sistemas dinámicos (multicuerpo). El filtrado generalizado proporciona densidades posteriores sobre estados ocultos (y parámetros) que generan datos observados utilizando un descenso de gradiente generalizado en energía libre variacional, bajo el . A diferencia del filtrado clásico (p. ej. Kalman-Bucy o de partículas), el filtrado generalizado evita las suposiciones markovianas sobre fluctuaciones aleatorias. Además, opera en (es) |
| rdfs:label | Filtrado generalizado (es) Generalized filtering (en) |
| owl:sameAs | freebase:Generalized filtering yago-res:Generalized filtering wikidata:Generalized filtering dbpedia-es:Generalized filtering https://global.dbpedia.org/id/fjEk |
| prov:wasDerivedFrom | wikipedia-en:Generalized_filtering?oldid=1069927937&ns=0 |
| foaf:isPrimaryTopicOf | wikipedia-en:Generalized_filtering |
| is dbo:wikiPageWikiLink of | dbr:Particle_filter dbr:Dynamic_Bayesian_network dbr:Dynamic_causal_modeling dbr:Andy_Clark dbr:Linear_predictive_coding dbr:Kalman_filter dbr:Karl_J._Friston dbr:Free_energy_principle dbr:Optimal_control dbr:System_identification dbr:Smoothing_problem_(stochastic_processes) dbr:Variational_Bayesian_methods dbr:Outline_of_machine_learning |
| is foaf:primaryTopic of | wikipedia-en:Generalized_filtering |