dc.contributor.author | Ashrafi, Parivash | |
dc.contributor.author | Sun, Yi | |
dc.contributor.author | Davey, Neil | |
dc.contributor.author | Adams, Roderick | |
dc.contributor.author | Wilkinson, Simon C | |
dc.contributor.author | Moss, Gary Patrick | |
dc.date.accessioned | 2018-04-10T18:38:26Z | |
dc.date.available | 2018-04-10T18:38:26Z | |
dc.date.issued | 2018-02-14 | |
dc.identifier.citation | Ashrafi , P , Sun , Y , Davey , N , Adams , R , Wilkinson , S C & Moss , G P 2018 , ' Model fitting for small skin permeability data sets: hyperparameter optimisation in Gaussian Process Regression ' , Journal of Pharmacy and Pharmacology , vol. 70 , no. 3 , pp. 361-373 . https://doi.org/10.1111/jphp.12863 | |
dc.identifier.issn | 0022-3573 | |
dc.identifier.uri | http://hdl.handle.net/2299/19964 | |
dc.description | This is the pre-peer reviewed version of the following article: Parivash Ashrafi, Yi Sun, Neil Davey, Roderick G. Adams, Simon C. Wilkinson, and Gary Patrick Moss, ‘Model fitting for small skin permeability data sets: hyperparameter optimisation in Gaussian Process Regression’, Journal of Pharmacy and Pharmacology, Vol. 70 (3): 361-373, March 2018, which has been published in final form at https://doi.org/10.1111/jphp.12863. Under embargo until 17 January 2019. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving. | |
dc.description.abstract | Objectives The aim of this study was to investigate how to improve predictions from Gaussian Process models by optimising the model hyperparameters. Methods Optimisation methods, including Grid Search, Conjugate Gradient, Random Search, Evolutionary Algorithm and Hyper-prior, were evaluated and applied to previously published data. Data sets were also altered in a structured manner to reduce their size, which retained the range, or ‘chemical space’ of the key descriptors to assess the effect of the data range on model quality. Key findings The Hyper-prior Smoothbox kernel results in the best models for the majority of data sets, and they exhibited significantly better performance than benchmark quantitative structure–permeability relationship (QSPR) models. When the data sets were systematically reduced in size, the different optimisation methods generally retained their statistical quality, whereas benchmark QSPR models performed poorly. Conclusions The design of the data set, and possibly also the approach to validation of the model, is critical in the development of improved models. The size of the data set, if carefully controlled, was not generally a significant factor for these models and that models of excellent statistical quality could be produced from substantially smaller data sets. | en |
dc.format.extent | 13 | |
dc.format.extent | 1517190 | |
dc.language.iso | eng | |
dc.relation.ispartof | Journal of Pharmacy and Pharmacology | |
dc.subject | Gaussian process | |
dc.subject | hyperparameters | |
dc.subject | machine learning | |
dc.subject | quantitative structure–permeability relationship | |
dc.subject | skin permeability | |
dc.subject | Pharmacology | |
dc.subject | Pharmaceutical Science | |
dc.title | Model fitting for small skin permeability data sets: hyperparameter optimisation in Gaussian Process Regression | en |
dc.contributor.institution | School of Computer Science | |
dc.contributor.institution | Centre for Computer Science and Informatics Research | |
dc.contributor.institution | Biocomputation Research Group | |
dc.description.status | Peer reviewed | |
dc.date.embargoedUntil | 2019-01-17 | |
dc.identifier.url | http://www.scopus.com/inward/record.url?scp=85040724399&partnerID=8YFLogxK | |
rioxxterms.versionofrecord | 10.1111/jphp.12863 | |
rioxxterms.type | Journal Article/Review | |
herts.preservation.rarelyaccessed | true | |