Comparing Accuracy Performance of ELM, ARMA and ARMA-GARCH Model In Predicting Exchange Rate Return

Alphanumeric Journal

MeldaAkın¹  Nimet Melis Esenyel İçen¹ 

[1] Istanbul University, Department of Econometrics, Faculty of Economics;

关键词: ARMA;    ARMA-GARCH;    Artificial Neural Networks;    Extreme Learning Machine;   

DOI  :  http://dx.doi.org/10.17093/alphanumeric.298658

来源: DOAJ

【 摘 要 】

GARCH type models and artificial intelligence models are frequently used in the modeling of financial time series returns. In this study, the performance of ARMA and ARMA-GARCH models was compared with ELM. Four error measurement criteria were used in the performance comparison. According to the findings, ELM models of Euro and GBP exchange rates returns are superior to the ARMA and ARMA-GARCH models. According to this result, it can be said that ELM, one of the artificial intelligence-based methods, is more suitable for estimating the exchange rate returns during the period covered.

【 授权许可】

Unknown