The theta model: a decomposition approach to forecasting

Assimakopoulos, Vassilios; Νικολόπουλος, Κωνσταντίνος

doi:10.1016/s0169-2070(00)00066-2

Cited by 375 publications

(247 citation statements)

References 7 publications

Supporting

Mentioning

218

Contrasting

Unclassified

Order By: Relevance

“…We implement a simple version of the theta forecasting method, which was introduced in [49], through the thetaf function of the R package forecast. Theta uses a given time series to create two or more auxiliary time series with different (modified) local curvatures with respect to the original, namely the "Theta-lines".…”

Section: Forecasting Using the Theta Methodsmentioning

confidence: 99%

“…The naïve methods usually perform well in predicting time series [37] and the ARFIMA processes are traditional methods which are frequently used for time series forecasting [45]. The theta method has been recently introduced and is also one of the most successful forecasting methods [49]. We use various versions of the RF algorithm with respect to the number of predictor variables, while the optimization of the parameter set is performed using methods proposed in [50,51].…”

Section: A Framework To Assess the Performance Of Random Forests In Tmentioning

confidence: 99%

See 1 more Smart Citation

Variable Selection in Time Series Forecasting Using Random Forests

2017

View full text Add to dashboard Cite

Time series forecasting using machine learning algorithms has gained popularity recently. Random forest is a machine learning algorithm implemented in time series forecasting; however, most of its forecasting properties have remained unexplored. Here we focus on assessing the performance of random forests in one-step forecasting using two large datasets of short time series with the aim to suggest an optimal set of predictor variables. Furthermore, we compare its performance to benchmarking methods. The first dataset is composed by 16,000 simulated time series from a variety of Autoregressive Fractionally Integrated Moving Average (ARFIMA) models. The second dataset consists of 135 mean annual temperature time series. The highest predictive performance of RF is observed when using a low number of recent lagged predictor variables. This outcome could be useful in relevant future applications, with the prospect to achieve higher predictive accuracy.

show abstract

Section: Forecasting Using the Theta Methodsmentioning

confidence: 99%

Section: A Framework To Assess the Performance Of Random Forests In Tmentioning

confidence: 99%

Variable Selection in Time Series Forecasting Using Random Forests

2017

View full text Add to dashboard Cite

show abstract

“…• Decomposition models: classical additive or multiplicative decomposition; Theta Model (Assimakopoulos & Nikolopoulos, 2000) • Exponential Smoothing Models: Simple smoothing; Holt (Holt et al, 1960); HoltWinter (additive or multiplicative) (Winters, 1960); Croston (Croston, 1972); Syntetos-Boylan (Syntetos & Boylan, 2005) • Moving Average Models: Simple moving average; Weighted simple moving average…”

Section: Application To a Real Casementioning

confidence: 99%

Forecasting model selection through out-of-sample rolling horizon weighted errors

Mula

2011

Expert Systems with Applications

View full text Add to dashboard Cite

ElsevierPoler Escoto, R.; Mula, J. (2011) . AbstractDemand Forecasting is an essential process for any firm whether it is a supplier, manufacturer or retailer. A large number of research works about time series forecast techniques exists in the literature, and there are many time series forecasting tools. In many cases, however, selecting the best time series forecasting model for each time series to be dealt with is still a complex problem. In this paper, a new automatic selection procedure of time series forecasting models is proposed. The selection criterion has been tested using the set of monthly time series of the M3 Competition and two basic forecasting models obtaining interesting results. This selection criterion has been implemented in a forecasting expert system and applied to a real case, a firm that produces steel products for construction, which automatically performs monthly forecasts on tens of thousands of time series. As result, the firm has increased the level of success in its demand forecasts.

show abstract

“…Através da série filtrada pelos dois métodos, a Echo State Networké treinada e a previsão 18-passos-a-frente são efetuadas. Ainda, observa-se a presença de modelos de previsão reconhecidos pela literatura científica, como o método Theta [17], assim como pela indústria (Autobox e ForecastPro) nas melhores colocações da competição. A crítica que pode ser estabelecida a tal competição reside no longo horizonte de previsão (18 passos-a-frente) e por somente conter séries cuja periodicidadeé mensal.…”

Section: Resultados E Discussõesunclassified

“…, 0} (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) o objetivo do P CD ké contabilizar o grau de ativação das demais classes, e descontá-las do CD k . Assim, estar-se-ia buscando regras que são mais específicas a uma determinada classe e não as mais generalistas.…”

Section: Máxima Confiança (Acurácia Fuzzy)unclassified

Gpfis: Um Sistema Fuzzy-Genético Genérico Baseado Em Programação Genética

KOSHIYAMA¹

View full text Add to dashboard Cite

AgradecimentosAo povo brasileiro, ao CNPq e a PUC-Rio pelos auxílios e estrutura concedida, sem os quais este trabalho não poderia ter sido realizado.A minha família, em destaque meus irmãos -Wladmir e Kallyah -e pais -Adriano e Sayonara -por todos os momentos abdicados da minha presença, devido a este manuscrito e demais projetos. Agradeço e saibam que temos uma ligação de afeto para além do infinito.A minha orientadora, Professora Marley Vellasco, que sempre me estimulou a produzir cientificamente, me incentivando nas inúmeras reuniões e revisões de texto, além da oportunidade de trabalhar em projetos. Também ao meu co-orientador, Professor Ricardo Tanscheit, pelas excelentes conversas e histórias compartilhadas e por ter me permitido absorver um pouco do seu método "minimalista". A ambos o meu maior respeito e admiração.Aos Também a pessoas com quem pude mais aprender do que transmitir, e com as quais formou-se uma amizade sólida. Iury Steiner, João Zanni, Pedro Milet e Douglas Dias, a vocês obrigado por todas as risadas e experiência passada. Ainda, a Tatiana Escovedo e Rubens Nascimento que me adotaram como filho e partilharam comigo trabalhos, pizzas e viagens.Aos professores Marco Pacheco, Wagner Tassinari,Álvaro Veiga, Carlos Tomei e Maria Lorenzon, dos quais pude absorver, cada um a sua maneira, a forma de lidar e transmitir a ciência, sempre com rigor e crítica, e me permitiram também vislumbrar a aplicação desta fora dos muros da universidade. In memorian a Padre Libanio pela amizade partilhada e exemplo de vida.Por fim, ao corpo de funcionários do Departamento de Engenharia Elétrica, notavelmente Alcina, Mara, Márcia, Isnarde e Miriam pelo apoio.

show abstract

The theta model: a decomposition approach to forecasting

Cited by 375 publications

References 7 publications

Variable Selection in Time Series Forecasting Using Random Forests

Variable Selection in Time Series Forecasting Using Random Forests

Forecasting model selection through out-of-sample rolling horizon weighted errors

Gpfis: Um Sistema Fuzzy-Genético Genérico Baseado Em Programação Genética

Contact Info

Product

Resources

About