Gaussian processes for global optimization
We introduce a novel Bayesian approach to global optimization using Gaussian processes. We frame the optimization of both noisy and noiseless functions as sequential decision problems, and introduce myopic and non-myopic solutions to them. Here our solutions can be tailored to exactly the degree of...
| Main Authors: | , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
2009
|
| _version_ | 1826309795482697728 |
|---|---|
| author | Osborne, MA Garnett, R Roberts, SJ |
| author_facet | Osborne, MA Garnett, R Roberts, SJ |
| author_sort | Osborne, MA |
| collection | OXFORD |
| description | We introduce a novel Bayesian approach to global optimization using Gaussian processes. We frame the optimization of both noisy and noiseless functions as sequential decision problems, and introduce myopic and non-myopic solutions to them. Here our solutions can be tailored to exactly the degree of confidence we require of them. The use of Gaussian processes allows us to benefit from the incorporation of prior knowledge about our objective function, and also from any derivative observations. Using this latter fact, we introduce an innovative method to combat conditioning problems. Our algorithm demonstrates a significant improvement over its competitors in overall performance across a wide range of canonical test problems. |
| first_indexed | 2024-03-07T07:41:03Z |
| format | Conference item |
| id | oxford-uuid:7d2b38d0-43be-4bb4-852c-50001a28ead9 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T07:41:03Z |
| publishDate | 2009 |
| record_format | dspace |
| spelling | oxford-uuid:7d2b38d0-43be-4bb4-852c-50001a28ead92023-04-28T08:32:35ZGaussian processes for global optimizationConference itemhttp://purl.org/coar/resource_type/c_5794uuid:7d2b38d0-43be-4bb4-852c-50001a28ead9EnglishSymplectic Elements2009Osborne, MAGarnett, RRoberts, SJWe introduce a novel Bayesian approach to global optimization using Gaussian processes. We frame the optimization of both noisy and noiseless functions as sequential decision problems, and introduce myopic and non-myopic solutions to them. Here our solutions can be tailored to exactly the degree of confidence we require of them. The use of Gaussian processes allows us to benefit from the incorporation of prior knowledge about our objective function, and also from any derivative observations. Using this latter fact, we introduce an innovative method to combat conditioning problems. Our algorithm demonstrates a significant improvement over its competitors in overall performance across a wide range of canonical test problems. |
| spellingShingle | Osborne, MA Garnett, R Roberts, SJ Gaussian processes for global optimization |
| title | Gaussian processes for global optimization |
| title_full | Gaussian processes for global optimization |
| title_fullStr | Gaussian processes for global optimization |
| title_full_unstemmed | Gaussian processes for global optimization |
| title_short | Gaussian processes for global optimization |
| title_sort | gaussian processes for global optimization |
| work_keys_str_mv | AT osbornema gaussianprocessesforglobaloptimization AT garnettr gaussianprocessesforglobaloptimization AT robertssj gaussianprocessesforglobaloptimization |