Prediction of foreign exchange market states with support vector machine

Kei Shioda, Shangkun Deng, Akito Sakurai

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

5 Citations (Scopus)

Abstract

This paper proposes a method to give an early warning of an abrupt change of price in a foreign exchange market. Volatility is a quantification of how much a value moves in a time series. It is now customary to assume that volatility of foreign exchange markets is time-varying. Intuitively we observe that there are at least two states or regimes: one is with low volatility and the other is with high volatility. Under high volatility regime, there are chances of high returns but with very high risks. For many nonprofessional traders, the high volatility regimes are periods that they loose with high probability. We believe that giving an early alert of starts of high volatility regimes is beneficial for many nonprofessional traders and for the foreign exchange markets. There are many studies to predict volatility of foreign exchange market by using ARCH or GARCH model with possibly hidden Markov model to represent regimes. We, though, focused on prediction of volatility levels by using machine learning techniques so that we get a good prediction. We particularly focused on support vector machine that learns sequences of volatility levels estimated by hidden Markov model and makes prediction of the level. We performed numerical experiments on real data and obtained good performance.

Original languageEnglish
Title of host publicationProceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011
Pages327-332
Number of pages6
Volume1
DOIs
Publication statusPublished - 2011
Event10th International Conference on Machine Learning and Applications, ICMLA 2011 - Honolulu, HI, United States
Duration: 2011 Dec 182011 Dec 21

Other

Other10th International Conference on Machine Learning and Applications, ICMLA 2011
CountryUnited States
CityHonolulu, HI
Period11/12/1811/12/21

Fingerprint

Support vector machines
Hidden Markov models
Learning systems
Time series
Financial markets
Experiments

Keywords

  • Foreign exchange
  • hidden markov model
  • machine learning
  • prediction
  • support vector machine
  • volatility

ASJC Scopus subject areas

  • Computer Science Applications
  • Human-Computer Interaction

Cite this

Shioda, K., Deng, S., & Sakurai, A. (2011). Prediction of foreign exchange market states with support vector machine. In Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011 (Vol. 1, pp. 327-332). [6146993] https://doi.org/10.1109/ICMLA.2011.116

Prediction of foreign exchange market states with support vector machine. / Shioda, Kei; Deng, Shangkun; Sakurai, Akito.

Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011. Vol. 1 2011. p. 327-332 6146993.

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

Shioda, K, Deng, S & Sakurai, A 2011, Prediction of foreign exchange market states with support vector machine. in Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011. vol. 1, 6146993, pp. 327-332, 10th International Conference on Machine Learning and Applications, ICMLA 2011, Honolulu, HI, United States, 11/12/18. https://doi.org/10.1109/ICMLA.2011.116
Shioda K, Deng S, Sakurai A. Prediction of foreign exchange market states with support vector machine. In Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011. Vol. 1. 2011. p. 327-332. 6146993 https://doi.org/10.1109/ICMLA.2011.116
Shioda, Kei ; Deng, Shangkun ; Sakurai, Akito. / Prediction of foreign exchange market states with support vector machine. Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011. Vol. 1 2011. pp. 327-332
@inproceedings{5e07718e587a43a193485b619774be51,
title = "Prediction of foreign exchange market states with support vector machine",
abstract = "This paper proposes a method to give an early warning of an abrupt change of price in a foreign exchange market. Volatility is a quantification of how much a value moves in a time series. It is now customary to assume that volatility of foreign exchange markets is time-varying. Intuitively we observe that there are at least two states or regimes: one is with low volatility and the other is with high volatility. Under high volatility regime, there are chances of high returns but with very high risks. For many nonprofessional traders, the high volatility regimes are periods that they loose with high probability. We believe that giving an early alert of starts of high volatility regimes is beneficial for many nonprofessional traders and for the foreign exchange markets. There are many studies to predict volatility of foreign exchange market by using ARCH or GARCH model with possibly hidden Markov model to represent regimes. We, though, focused on prediction of volatility levels by using machine learning techniques so that we get a good prediction. We particularly focused on support vector machine that learns sequences of volatility levels estimated by hidden Markov model and makes prediction of the level. We performed numerical experiments on real data and obtained good performance.",
keywords = "Foreign exchange, hidden markov model, machine learning, prediction, support vector machine, volatility",
author = "Kei Shioda and Shangkun Deng and Akito Sakurai",
year = "2011",
doi = "10.1109/ICMLA.2011.116",
language = "English",
isbn = "9780769546070",
volume = "1",
pages = "327--332",
booktitle = "Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011",

}

TY - GEN

T1 - Prediction of foreign exchange market states with support vector machine

AU - Shioda, Kei

AU - Deng, Shangkun

AU - Sakurai, Akito

PY - 2011

Y1 - 2011

N2 - This paper proposes a method to give an early warning of an abrupt change of price in a foreign exchange market. Volatility is a quantification of how much a value moves in a time series. It is now customary to assume that volatility of foreign exchange markets is time-varying. Intuitively we observe that there are at least two states or regimes: one is with low volatility and the other is with high volatility. Under high volatility regime, there are chances of high returns but with very high risks. For many nonprofessional traders, the high volatility regimes are periods that they loose with high probability. We believe that giving an early alert of starts of high volatility regimes is beneficial for many nonprofessional traders and for the foreign exchange markets. There are many studies to predict volatility of foreign exchange market by using ARCH or GARCH model with possibly hidden Markov model to represent regimes. We, though, focused on prediction of volatility levels by using machine learning techniques so that we get a good prediction. We particularly focused on support vector machine that learns sequences of volatility levels estimated by hidden Markov model and makes prediction of the level. We performed numerical experiments on real data and obtained good performance.

AB - This paper proposes a method to give an early warning of an abrupt change of price in a foreign exchange market. Volatility is a quantification of how much a value moves in a time series. It is now customary to assume that volatility of foreign exchange markets is time-varying. Intuitively we observe that there are at least two states or regimes: one is with low volatility and the other is with high volatility. Under high volatility regime, there are chances of high returns but with very high risks. For many nonprofessional traders, the high volatility regimes are periods that they loose with high probability. We believe that giving an early alert of starts of high volatility regimes is beneficial for many nonprofessional traders and for the foreign exchange markets. There are many studies to predict volatility of foreign exchange market by using ARCH or GARCH model with possibly hidden Markov model to represent regimes. We, though, focused on prediction of volatility levels by using machine learning techniques so that we get a good prediction. We particularly focused on support vector machine that learns sequences of volatility levels estimated by hidden Markov model and makes prediction of the level. We performed numerical experiments on real data and obtained good performance.

KW - Foreign exchange

KW - hidden markov model

KW - machine learning

KW - prediction

KW - support vector machine

KW - volatility

UR - http://www.scopus.com/inward/record.url?scp=84857815714&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84857815714&partnerID=8YFLogxK

U2 - 10.1109/ICMLA.2011.116

DO - 10.1109/ICMLA.2011.116

M3 - Conference contribution

SN - 9780769546070

VL - 1

SP - 327

EP - 332

BT - Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011

ER -