{"title":"Streamlining hyperparameter optimization for radiation emulator training with automated Sherpa","authors":"Soonyoung Roh, Park Sa Kim, Hwan-Jin Song","doi":"10.1186/s40562-024-00336-8","DOIUrl":null,"url":null,"abstract":"This study aimed to identify the optimal configuration for neural network (NN) emulators in numerical weather prediction, minimizing trial and error by comparing emulator performance across multiple hidden layers (1–5 layers), as automatically defined by the Sherpa library. Our findings revealed that Sherpa-applied emulators consistently demonstrated good results and stable performance with low errors in numerical simulations. The optimal configurations were observed with one and two hidden layers, improving results when two hidden layers were employed. The Sherpa-defined average neurons per hidden layer ranged between 153 and 440, resulting in a speedup relative to the CNT of 7–12 times. These results provide valuable insights for developing radiative physical NN emulators. Utilizing automatically determined hyperparameters can effectively reduce trial-and-error processes while maintaining stable outcomes. However, further experimentation is needed to establish the most suitable hyperparameter values that balance both speed and accuracy, as this study did not identify optimized values for all hyperparameters. ","PeriodicalId":48596,"journal":{"name":"Geoscience Letters","volume":"61 1","pages":""},"PeriodicalIF":4.0000,"publicationDate":"2024-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Geoscience Letters","FirstCategoryId":"89","ListUrlMain":"https://doi.org/10.1186/s40562-024-00336-8","RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GEOSCIENCES, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
This study aimed to identify the optimal configuration for neural network (NN) emulators in numerical weather prediction, minimizing trial and error by comparing emulator performance across multiple hidden layers (1–5 layers), as automatically defined by the Sherpa library. Our findings revealed that Sherpa-applied emulators consistently demonstrated good results and stable performance with low errors in numerical simulations. The optimal configurations were observed with one and two hidden layers, improving results when two hidden layers were employed. The Sherpa-defined average neurons per hidden layer ranged between 153 and 440, resulting in a speedup relative to the CNT of 7–12 times. These results provide valuable insights for developing radiative physical NN emulators. Utilizing automatically determined hyperparameters can effectively reduce trial-and-error processes while maintaining stable outcomes. However, further experimentation is needed to establish the most suitable hyperparameter values that balance both speed and accuracy, as this study did not identify optimized values for all hyperparameters.
Geoscience LettersEarth and Planetary Sciences-General Earth and Planetary Sciences
CiteScore
4.90
自引率
2.50%
发文量
42
审稿时长
25 weeks
期刊介绍:
Geoscience Letters is the official journal of the Asia Oceania Geosciences Society, and a fully open access journal published under the SpringerOpen brand. The journal publishes original, innovative and timely research letter articles and concise reviews on studies of the Earth and its environment, the planetary and space sciences. Contributions reflect the eight scientific sections of the AOGS: Atmospheric Sciences, Biogeosciences, Hydrological Sciences, Interdisciplinary Geosciences, Ocean Sciences, Planetary Sciences, Solar and Terrestrial Sciences, and Solid Earth Sciences. Geoscience Letters focuses on cutting-edge fundamental and applied research in the broad field of the geosciences, including the applications of geoscience research to societal problems. This journal is Open Access, providing rapid electronic publication of high-quality, peer-reviewed scientific contributions.