Foreign exchange prediction and trading using long-short-term-memory neural network

Foreign exchange (forex) market is the largest financial market in the world. The relevance of this market results in many researches carried out to study its operation. In the past few years, many techniques have been developed to study, including technical analysis, where researchers tried to pred...

Full description

Saved in:
Bibliographic Details
Main Author: Tan, Shyer Bin
Other Authors: Wang Lipo
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2020
Subjects:
Online Access:https://hdl.handle.net/10356/139217
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Foreign exchange (forex) market is the largest financial market in the world. The relevance of this market results in many researches carried out to study its operation. In the past few years, many techniques have been developed to study, including technical analysis, where researchers tried to predict the forex rate based on past forex data. Neural network is a tool for predicting value based on training data. In the context of forex prediction, the past trading data is used as training data. The sheer size of trading volume means good amount of data for time series analysis and prediction. Researchers try to explore the intrinsic relationships in trading data, understand the pattern of fluctuation of the forex rate, and make use of these models to build accurate models to predict future rates. A signal decomposition method will also be used to reduce the complexity of the highly liquid forex data. In this project, the work done by published research papers will first be replicated. This is to set up a standard for comparison for different predicting models, using same set of data and comparable parameters for the neural networks. Next, the focus of the project will be improving the performance of prediction, where Long-Short-Term Memory Neural Network (LSTM) and Variational Mode Decomposition (VMD) technique are used to predict the forex rate.