File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.26802/jaots.2018.09452
- Scopus: eid_2-s2.0-85051342901
- Find via
Supplementary
-
Citations:
- Scopus: 0
- Appears in Collections:
Article: Gaussian process based ITO evolution algorithm to avoid local minima in machine learning and cloud computing
Title | Gaussian process based ITO evolution algorithm to avoid local minima in machine learning and cloud computing |
---|---|
Authors | |
Keywords | Fluctuation Ratio Category Theory Gaussian Process Incremental Inheritance ITO |
Issue Date | 2018 |
Citation | Journal of Advanced Oxidation Technologies, 2018, v. 21, n. 2, article no. 201809452 How to Cite? |
Abstract | © 2018 Walter de Gruyter GmbH. All rights reserved. We introduce the Gaussian process (GP) regression model as ITO's fluctuation operator in mobility model, to help ITO get out of possible local minima. To enhance the global search and local search link as a starting point, an ITO algorithm based on Gaussian process model is put forward. In this paper, to enhance the global search and local search link as a starting point, an ITO algorithm based on Gaussian process model is put forward. The equivalence between the Gaussian process model and wave operator of the ITO algorithm are proved. By GP-based particle movement mode in the fluctuations process of local search, particles' search behaviors are guided with the ability to jump out of local minimum. Besides, the convergence speed of particles is also enhanced. The final experimental results show that the GITO algorithm proposed in this paper has better convergence speed and solution performance. This paper preliminary established the "link" between ITO algorithm local search and global search by using machine-learning method based on Gaussian process. To illustrate this, kernel function of Gaussian process model, which is relatively simple, does not take into account the special scene of wave process. Under the guide of this operator, ITO's capacity of local searching and global searching is strengthened to some extent. Finally, Comparing with origin algorithm, the experiments show our improved ITO algorithm's better convergence rate and performance. Gaussian process ITO fluctuation ratio. |
Persistent Identifier | http://hdl.handle.net/10722/296178 |
ISSN | 2017 Impact Factor: 0.901 2019 SCImago Journal Rankings: 0.106 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Ma, Chuang | - |
dc.contributor.author | Yang, Yongjian | - |
dc.contributor.author | Du, Zhanwei | - |
dc.date.accessioned | 2021-02-11T04:53:00Z | - |
dc.date.available | 2021-02-11T04:53:00Z | - |
dc.date.issued | 2018 | - |
dc.identifier.citation | Journal of Advanced Oxidation Technologies, 2018, v. 21, n. 2, article no. 201809452 | - |
dc.identifier.issn | 1203-8407 | - |
dc.identifier.uri | http://hdl.handle.net/10722/296178 | - |
dc.description.abstract | © 2018 Walter de Gruyter GmbH. All rights reserved. We introduce the Gaussian process (GP) regression model as ITO's fluctuation operator in mobility model, to help ITO get out of possible local minima. To enhance the global search and local search link as a starting point, an ITO algorithm based on Gaussian process model is put forward. In this paper, to enhance the global search and local search link as a starting point, an ITO algorithm based on Gaussian process model is put forward. The equivalence between the Gaussian process model and wave operator of the ITO algorithm are proved. By GP-based particle movement mode in the fluctuations process of local search, particles' search behaviors are guided with the ability to jump out of local minimum. Besides, the convergence speed of particles is also enhanced. The final experimental results show that the GITO algorithm proposed in this paper has better convergence speed and solution performance. This paper preliminary established the "link" between ITO algorithm local search and global search by using machine-learning method based on Gaussian process. To illustrate this, kernel function of Gaussian process model, which is relatively simple, does not take into account the special scene of wave process. Under the guide of this operator, ITO's capacity of local searching and global searching is strengthened to some extent. Finally, Comparing with origin algorithm, the experiments show our improved ITO algorithm's better convergence rate and performance. Gaussian process ITO fluctuation ratio. | - |
dc.language | eng | - |
dc.relation.ispartof | Journal of Advanced Oxidation Technologies | - |
dc.subject | Fluctuation Ratio | - |
dc.subject | Category Theory | - |
dc.subject | Gaussian Process | - |
dc.subject | Incremental Inheritance | - |
dc.subject | ITO | - |
dc.title | Gaussian process based ITO evolution algorithm to avoid local minima in machine learning and cloud computing | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.26802/jaots.2018.09452 | - |
dc.identifier.scopus | eid_2-s2.0-85051342901 | - |
dc.identifier.volume | 21 | - |
dc.identifier.issue | 2 | - |
dc.identifier.spage | article no. 201809452 | - |
dc.identifier.epage | article no. 201809452 | - |
dc.identifier.eissn | 2371-1175 | - |
dc.identifier.issnl | 1203-8407 | - |