Robust forex trading with Deep Q Network (DQN)
Robust forex trading with Deep Q Network (DQN)
Files (excerpt)
Published date
2019
Resource type
Publisher
Bangkok : Assumption University
ISBN
ISSN
DOI
Call no.
Other identifier(s)
Edition
Copyrighted date
Language
eng
File type
application/pdf
Extent
19 pages
Other title(s)
Authors
Advisor
Other Contributor(s)
Citation
ABAC Journal Vol. 39 No.1 (January-March 2019), 15-33
Degree name
Degree level
Degree discipline
Degree department
Degree grantor
Abstract
Financial trading is one of the most attractive areas in finance. Trading
systems development is not an easy task because it requires extensive
knowledge in several areas such as quantitative analysis, financial skills, and
computer programming. A trading systems expert, as a human, also brings in
their own bias when developing the system. There should be another, more
effective way to develop the system using artificial intelligence. The aim of
this study was to compare the performance of AI agents to the performance of
the buy-and-hold strategy and the expert trader. The tested market consisted
of 15 years of the Forex data market, from two currency pairs (EURUSD,
USDJPY) obtained from Dukascopy Bank SA Switzerland. Both hypotheses
were tested with a paired t-Test at the 0.05 significance level. The findings
showed that AI can beat the buy & hold strategy with significant superiority,
in FOREX for both currency pairs (EURUSD, USDJPY), and that AI can also
significantly outperform CTA (experienced trader) for trading in EURUSD.
However, the AI could not significantly outperform CTA for USDJPY trading.
Limitations, contributions, and further research were recommended.
Table of contents
Description
punsarn.dc.description.sponsorship
Spatial Coverage
Subject(s)
Rights
This work is protected by copyright. Reproduction or distribution of the work in any format is prohibited without written permission of the copyright owner.