Bayesian prediction via partitioning
This article proposes a new Bayesian approach to prediction on continuous covariates. The Bayesian partition model constructs arbitrarily complex regression and classification surfaces by splitting the covariate space into an unknown number of disjoint regions. Within each region the data are assume...
| Main Authors: | , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
2005
|
| _version_ | 1826258876497920000 |
|---|---|
| author | Holmes, C Denison, D Ray, S Mallick, B |
| author_facet | Holmes, C Denison, D Ray, S Mallick, B |
| author_sort | Holmes, C |
| collection | OXFORD |
| description | This article proposes a new Bayesian approach to prediction on continuous covariates. The Bayesian partition model constructs arbitrarily complex regression and classification surfaces by splitting the covariate space into an unknown number of disjoint regions. Within each region the data are assumed to be exchangeable and come from some simple distribution. Using conjugate priors, the marginal likelihoods of the models can be obtained analytically for any proposed partitioning of the space where the number and location of the regions is assumed unknown a priori. Markov chain Monte Carlo simulation techniques are used to obtain predictive distributions at the design points by averaging across posterior samples of partitions. © 2005 American Statistical Association, Institute of Mathematical Statistics, and Interface Foundation of North America. |
| first_indexed | 2024-03-06T18:40:57Z |
| format | Journal article |
| id | oxford-uuid:0cd7f6cc-87f9-4a54-b1c1-cf179a1ef7c6 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T18:40:57Z |
| publishDate | 2005 |
| record_format | dspace |
| spelling | oxford-uuid:0cd7f6cc-87f9-4a54-b1c1-cf179a1ef7c62022-03-26T09:37:17ZBayesian prediction via partitioningJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:0cd7f6cc-87f9-4a54-b1c1-cf179a1ef7c6EnglishSymplectic Elements at Oxford2005Holmes, CDenison, DRay, SMallick, BThis article proposes a new Bayesian approach to prediction on continuous covariates. The Bayesian partition model constructs arbitrarily complex regression and classification surfaces by splitting the covariate space into an unknown number of disjoint regions. Within each region the data are assumed to be exchangeable and come from some simple distribution. Using conjugate priors, the marginal likelihoods of the models can be obtained analytically for any proposed partitioning of the space where the number and location of the regions is assumed unknown a priori. Markov chain Monte Carlo simulation techniques are used to obtain predictive distributions at the design points by averaging across posterior samples of partitions. © 2005 American Statistical Association, Institute of Mathematical Statistics, and Interface Foundation of North America. |
| spellingShingle | Holmes, C Denison, D Ray, S Mallick, B Bayesian prediction via partitioning |
| title | Bayesian prediction via partitioning |
| title_full | Bayesian prediction via partitioning |
| title_fullStr | Bayesian prediction via partitioning |
| title_full_unstemmed | Bayesian prediction via partitioning |
| title_short | Bayesian prediction via partitioning |
| title_sort | bayesian prediction via partitioning |
| work_keys_str_mv | AT holmesc bayesianpredictionviapartitioning AT denisond bayesianpredictionviapartitioning AT rays bayesianpredictionviapartitioning AT mallickb bayesianpredictionviapartitioning |