Towards Optimal Supercomputer Energy Consumption Forecasting Method

Accurate prediction methods are generally very computationally intensive, so they take a long time. Quick prediction methods, on the other hand, are not very accurate. Is it possible to design a prediction method that is both accurate and fast? In this paper, a new prediction method is proposed, bas...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autor principal:	Jiří Tomčala
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	forecasting prediction method time series random time delays patterns zeroth algorithm machine learning Mathematics QA1-939
Acceso en línea:	https://doaj.org/article/5ee32ce0d4fe482492d1e8f7ac0030f3
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:5ee32ce0d4fe482492d1e8f7ac0030f3
record_format	dspace
spelling	oai:doaj.org-article:5ee32ce0d4fe482492d1e8f7ac0030f32021-11-11T18:15:36ZTowards Optimal Supercomputer Energy Consumption Forecasting Method10.3390/math92126952227-7390https://doaj.org/article/5ee32ce0d4fe482492d1e8f7ac0030f32021-10-01T00:00:00Zhttps://www.mdpi.com/2227-7390/9/21/2695https://doaj.org/toc/2227-7390Accurate prediction methods are generally very computationally intensive, so they take a long time. Quick prediction methods, on the other hand, are not very accurate. Is it possible to design a prediction method that is both accurate and fast? In this paper, a new prediction method is proposed, based on the so-called random time-delay patterns, named the RTDP method. Using these random time-delay patterns, this method looks for the most important parts of the time series’ previous evolution, and uses them to predict its future development. When comparing the supercomputer infrastructure power consumption prediction with other commonly used prediction methods, this newly proposed RTDP method proved to be the most accurate and the second fastest.Jiří TomčalaMDPI AGarticleforecastingprediction methodtime seriesrandom time delays patternszeroth algorithmmachine learningMathematicsQA1-939ENMathematics, Vol 9, Iss 2695, p 2695 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	forecasting prediction method time series random time delays patterns zeroth algorithm machine learning Mathematics QA1-939
spellingShingle	forecasting prediction method time series random time delays patterns zeroth algorithm machine learning Mathematics QA1-939 Jiří Tomčala Towards Optimal Supercomputer Energy Consumption Forecasting Method
description	Accurate prediction methods are generally very computationally intensive, so they take a long time. Quick prediction methods, on the other hand, are not very accurate. Is it possible to design a prediction method that is both accurate and fast? In this paper, a new prediction method is proposed, based on the so-called random time-delay patterns, named the RTDP method. Using these random time-delay patterns, this method looks for the most important parts of the time series’ previous evolution, and uses them to predict its future development. When comparing the supercomputer infrastructure power consumption prediction with other commonly used prediction methods, this newly proposed RTDP method proved to be the most accurate and the second fastest.
format	article
author	Jiří Tomčala
author_facet	Jiří Tomčala
author_sort	Jiří Tomčala
title	Towards Optimal Supercomputer Energy Consumption Forecasting Method
title_short	Towards Optimal Supercomputer Energy Consumption Forecasting Method
title_full	Towards Optimal Supercomputer Energy Consumption Forecasting Method
title_fullStr	Towards Optimal Supercomputer Energy Consumption Forecasting Method
title_full_unstemmed	Towards Optimal Supercomputer Energy Consumption Forecasting Method
title_sort	towards optimal supercomputer energy consumption forecasting method
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/5ee32ce0d4fe482492d1e8f7ac0030f3
work_keys_str_mv	AT jiritomcala towardsoptimalsupercomputerenergyconsumptionforecastingmethod
_version_	1718431888046030848

Towards Optimal Supercomputer Energy Consumption Forecasting Method

Ejemplares similares