Soft-computing techniques and ARMA model for time series prediction

Rojas, Ignacio; Valenzuela, Olga; Rojas, Fernando; Guillén, Alberto; Herrera, José Benavente; Pomares, H.; Marquez, L.; Pasadas, M.

doi:10.1016/j.neucom.2007.07.018

Cited by 199 publications

(81 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As shown in Fig.1 (2) In the input layer, the input variables comprise one input for the bias, the output vector C of the non-linear functional expansion block, three popular technical indicators, and one input for the recurrent link obtained by feeding back one step delayed output. Thus the final input vector is obtained as…”

Section: Development Of a Stock Forecasting System Using Pflarnn Modelmentioning

confidence: 99%

“…A significant amount of studies has been done in this field that includes hybrid combinations of soft computing technology and data mining analysis applied to stock data prediction over a time frame varying from one day ahead to several days ahead. Traditionally, the linear statistical models like autoregressive moving average (ARMA) or ARIM [1][2][3] used for time series forecasting are simple but suffer International Journal of Computational Intelligence Systems, Vol. 8, No.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Polynomial Based Functional Link Artificial Recurrent Neural Network adaptive System for predicting Indian Stocks

Bebarta¹,

Biswal²,

Dash³

2015

IJCIS

View full text Add to dashboard Cite

A low complexity Polynomial Functional link Artificial Recurrent Neural Network (PFLARNN) has been proposed for the prediction of financial time series data. Although different types of polynomial functions have been used for low complexity neural network architectures earlier for stock market prediction, a comparative study is needed to choose the optimal combinations of the nonlinear functions for a reasonably accurate forecast. Further a recurrent version of the Functional link neural network is used to model more accurately a chaotic time series like stock market indices with a lesser number of nonlinear basis functions. The proposed PFLARNN model when trained with the well known gradient descent algorithm produces reasonable accuracy with a choice of range of weight parameters of the network. However, to improve the accuracy of the forecast further, the weight parameters of the recurrent functional neural network are optimized using an evolutionary learning algorithm like the differential evolution (DE). A comparison with other well known neural architectures shows that the proposed low complexity neural model can provide significant prediction accuracy for one day advance and speed of convergence using the International Business Machines Corp. (IBM) stock market indices.

show abstract

Section: Development Of a Stock Forecasting System Using Pflarnn Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Polynomial Based Functional Link Artificial Recurrent Neural Network adaptive System for predicting Indian Stocks

Bebarta¹,

Biswal²,

Dash³

2015

IJCIS

View full text Add to dashboard Cite

show abstract

“…A hybrid of ARIMA and support vector machines was successfully presented by Pai et.al [8] for predicting stock prices problems. Other outstanding hybrid approaches could be found in references [9][10][11]. Most of these hybrid models were implemented as a following process: first, the model-based technique was used to predict the linear relation, then the data-driven based technique was utilized to forecast the residuals between actual values and predicted results obtained from previous step.…”

Section: Introductionmentioning

confidence: 99%

A hybrid of nonlinear autoregressive model with exogenous input and autoregressive moving average model for long-term machine state forecasting

Pham

Tran

Yang

2010

Expert Systems with Applications

View full text Add to dashboard Cite

This paper presents an improvement of hybrid of nonlinear autoregressive with exogenous input (NARX) and autoregressive moving average (ARMA) for long-term machine state forecasting based on vibration data. In this study, vibration data is considered as a combination of two components which are deterministic data and error. The deterministic component may describe the degradation index of machine, whilst the error component can depict the appearance of uncertain parts. An improved hybrid forecasting model, namely NARX-ARMA model, is carried out to obtain the forecasting results in which NARX network model which is suitable for nonlinear issue is used to forecast the deterministic component and ARMA model are used to predict the error component due to appropriate capability in linear prediction. The final forecasting results are the sum of the results obtained from these single models. The performance of the NARX-ARMA model is then evaluated by using the data of low methane compressor acquired from condition monitoring routine. In order to corroborate the advances of the proposed method, a comparative study of the forecasting results obtained from NARX-ARMA model and traditional models is also carried out. The comparative results show that NARX-ARMA model is outstanding and could be used as a potential tool to machine state forecasting.

show abstract

Section: Introductionunclassified

“…Huang e colaboradores descreveram em seu artigo a utilização de redes neurais artificiais na previsão de preços em uma série temporal financeira e as vantagens em se acoplar outros métodos de análise de séries temporais [5]. Rojas e colaboradores estudaram uma metodologia híbrida acoplando redes neurais artificiais e o modelo de séries temporais ARMA (Autoregressive Moving Average Model) [6].É nesse contexto, que o presente trabalho propõe um método de parametrização, por meio de algoritmo genético, do indicador de análise técnica do mercado financeiro MACD (Moving Average Convergence-Divergence) e a utilização de um sistema de inferência fuzzy que possa maximizar o lucro de investimentos nas ordens de compra e venda desse indicador. O…”

unclassified

Análise de Lucros em Investimentos a partir do MACD com o Emprego de Algoritmos Genéticos e Lógica Fuzzy

Marques¹,

Gomes²,

Almeida³

2009

L&NLM

View full text Add to dashboard Cite

-A new methodology for the parameterization of the technical analysis of the financial market indicator coinedMoving Average Convergence-Divergence (MACD) is presented in this paper. The architecture of the MACD involves the use of exponential moving averages that in turn use different time windows, tracking securities prices trends and signalising the right moment to purchase and sale shares. By using genetic algorithms, it was possible to establish an optimal value for the time window which could yield higher profits, when compared to the time window used in literature. The use of fuzzy logic indicates the best moment for purchasing and sale of shares, raising the security of each transaction, thus resulting in the increase of the success rate. The methodology proposed was validated by taking into account the Petrobras shares (PETR4) in the period between February 2005 and August 2008, achieving a profit higher than that in the usual parameterisation.Keywords -Genetic Algorithm, Fuzzy Logic, Financial Time Series. Resumo -Uma nova metodologia de parametrização do indicador de análise técnica do mercado financeiro chamadoMoving Average Convergence-Divergence (MACD) é apresentada neste artigo. A composição do MACD envolve o uso de médias móveis exponenciais que, por sua vez, utilizam janelas temporais diferentes, acompanhando a tendência dos preços dos valores mobiliários e indicando o melhor momento de compra e venda. Com o uso da técnica de algoritmos genéticos foi possível a escolha de janelas temporais que gerassem melhores lucros, quando comparados às janelas temporais utilizadas em diversas literaturas. A utilização de lógica fuzzy possibilitou a classificação das ordens de compra elevando a segurança de cada operação, traduzida no aumento da taxa de acerto. Para o trabalho aqui proposto entende-se com taxa de acerto a percentagem de operações de compra e venda que geraram rentabilidade positiva. A metodologia proposta foi validada utilizando as ações da Petrobras PETR4 no período entre novembro de 2006 e agosto de 2008, alcançando um lucro superior a parametrização usual.Palavras-chave -Algoritmo Genético, Lógica Fuzzy, Séries Temporais Financeiras. IntroduçãoA predição de dados em séries temporais é de extrema relevância para diversas áreas do conhecimento científico e vem sendo estudada desde o início do século XX quando alguns modelos matemáticos foram empregados. Até 1920, a predição de dados em séries temporais era basicamente realizada através da extrapolação dos dados por meio de um ajuste global no domínio do tempo. Entretanto, somente a partir de 1926 é que iniciou-se o que poderíamos chamar de moderna predição de séries, quando a técnica auto-regressiva foi publicada por Yule [1].Na área financeira, a análise de séries temporais tem grande importância no estudo dos fenômenos sócio-econômicos [2] nos quais não se é possível determinar a dinâmica de evolução dos processos envolvidos em sua composição.Muitos estudos têm sido desenvolvidos nesta área. Chen e colaboradores propuseram um mode...

show abstract

Soft-computing techniques and ARMA model for time series prediction

Cited by 199 publications

References 36 publications

Polynomial Based Functional Link Artificial Recurrent Neural Network adaptive System for predicting Indian Stocks

Polynomial Based Functional Link Artificial Recurrent Neural Network adaptive System for predicting Indian Stocks

A hybrid of nonlinear autoregressive model with exogenous input and autoregressive moving average model for long-term machine state forecasting

Análise de Lucros em Investimentos a partir do MACD com o Emprego de Algoritmos Genéticos e Lógica Fuzzy

Contact Info

Product

Resources

About