Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate

Shangkun Deng, Akito Sakurai

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

This paper proposes a hybrid model named MKL-GA, which combines Multiple Kernel Learning (MKL) and Genetic Algorithm (GA), for modeling and the prediction of FX (foreign exchange) rate on USDJPY currency pair by extracting features from three main FX pairs with three different short time horizons. Firstly, the MKL regression model predicts the change rate based on MACD indicators, and then GA is applied to fuse all the information from the regression model and overbought/oversold technical indicators. Experimental results show that the proposed model outperforms other models in terms of returns and risk-return ratio. In addition, the result of kernel weights for different currency pairs in the step of MKL training should be also advisable for the trading.

Original languageEnglish
Title of host publicationProceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011
Pages200-209
Number of pages10
DOIs
Publication statusPublished - 2011
Event11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011 - Innsbruck, Austria
Duration: 2011 Feb 142011 Feb 16

Other

Other11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011
CountryAustria
CityInnsbruck
Period11/2/1411/2/16

Fingerprint

Learning algorithms
Genetic algorithms
Electric fuses

Keywords

  • FX trading
  • Genetic Algorithm
  • MKL-GA hybrid model
  • Multiple Kernel Learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications

Cite this

Deng, S., & Sakurai, A. (2011). Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate. In Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011 (pp. 200-209) https://doi.org/10.2316/P.2011.717-111

Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate. / Deng, Shangkun; Sakurai, Akito.

Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011. 2011. p. 200-209.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Deng, S & Sakurai, A 2011, Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate. in Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011. pp. 200-209, 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011, Innsbruck, Austria, 11/2/14. https://doi.org/10.2316/P.2011.717-111
Deng S, Sakurai A. Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate. In Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011. 2011. p. 200-209 https://doi.org/10.2316/P.2011.717-111
Deng, Shangkun ; Sakurai, Akito. / Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate. Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011. 2011. pp. 200-209
@inproceedings{05686cf8094445608377f8e727ab7d1b,
title = "Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate",
abstract = "This paper proposes a hybrid model named MKL-GA, which combines Multiple Kernel Learning (MKL) and Genetic Algorithm (GA), for modeling and the prediction of FX (foreign exchange) rate on USDJPY currency pair by extracting features from three main FX pairs with three different short time horizons. Firstly, the MKL regression model predicts the change rate based on MACD indicators, and then GA is applied to fuse all the information from the regression model and overbought/oversold technical indicators. Experimental results show that the proposed model outperforms other models in terms of returns and risk-return ratio. In addition, the result of kernel weights for different currency pairs in the step of MKL training should be also advisable for the trading.",
keywords = "FX trading, Genetic Algorithm, MKL-GA hybrid model, Multiple Kernel Learning",
author = "Shangkun Deng and Akito Sakurai",
year = "2011",
doi = "10.2316/P.2011.717-111",
language = "English",
isbn = "9780889868632",
pages = "200--209",
booktitle = "Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011",

}

TY - GEN

T1 - Combining Multiple Kernel Learning and Genetic Algorithm for forecasting short time foreign exchange rate

AU - Deng, Shangkun

AU - Sakurai, Akito

PY - 2011

Y1 - 2011

N2 - This paper proposes a hybrid model named MKL-GA, which combines Multiple Kernel Learning (MKL) and Genetic Algorithm (GA), for modeling and the prediction of FX (foreign exchange) rate on USDJPY currency pair by extracting features from three main FX pairs with three different short time horizons. Firstly, the MKL regression model predicts the change rate based on MACD indicators, and then GA is applied to fuse all the information from the regression model and overbought/oversold technical indicators. Experimental results show that the proposed model outperforms other models in terms of returns and risk-return ratio. In addition, the result of kernel weights for different currency pairs in the step of MKL training should be also advisable for the trading.

AB - This paper proposes a hybrid model named MKL-GA, which combines Multiple Kernel Learning (MKL) and Genetic Algorithm (GA), for modeling and the prediction of FX (foreign exchange) rate on USDJPY currency pair by extracting features from three main FX pairs with three different short time horizons. Firstly, the MKL regression model predicts the change rate based on MACD indicators, and then GA is applied to fuse all the information from the regression model and overbought/oversold technical indicators. Experimental results show that the proposed model outperforms other models in terms of returns and risk-return ratio. In addition, the result of kernel weights for different currency pairs in the step of MKL training should be also advisable for the trading.

KW - FX trading

KW - Genetic Algorithm

KW - MKL-GA hybrid model

KW - Multiple Kernel Learning

UR - http://www.scopus.com/inward/record.url?scp=79958164800&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79958164800&partnerID=8YFLogxK

U2 - 10.2316/P.2011.717-111

DO - 10.2316/P.2011.717-111

M3 - Conference contribution

SN - 9780889868632

SP - 200

EP - 209

BT - Proceedings of the 11th IASTED International Conference on Artificial Intelligence and Applications, AIA 2011

ER -