Magdalene Ho Wen Lin
Traditionally, intuition of foreign currency traders or quantitative variables have been used to forecast the movement of foreign currency exchange (Forex) market for profit return. In recent years, sentiment analysis is gaining popularity in Forex forecasting and it has been shown to achieve a high accuracy in prediction in research literature. In this thesis, we build on the use of sentiment analysis for Forex prediction by incorporating the factor that causes the fluctuation of the price that may be due to the ambiguity of the world economy especially in the year of the widely spread pandemic. A Support Vector Machine (SVM) algorithm is used as the approach to train the model with the input of sentiment analysis. The collection of news headlines from different news sources is used to generate the sentiment value which is modelled together with the Forex movement through SVM in the creation of a real-world model for traders. It is expected that the sentiment of the news will have great influence on the Forex movement based on the promising past research. This research is significant for traders in the future where the world economy uncertainties would be more persistent in the coming years.
Magdalene Ho Wen Lin
This research focuses on the use of Sentiment Analysis to predict the movement of the Forex Currency
Sentiment analysis is an idea of extracting opinions and attitude from a language be it in a text or audio form, using various methods.
Foreign Exchange (Forex) trading is the buying and selling of the foreign currency in an international market such as exchanging US dollars for the UK pounds.
Hover to find out!
In the trading market, the aim is to make profit by predicting the movement of the price where we sell when the price is high and vice versa.
.
.
.
From the sentiment of the news headlines, machine learning will be able to accurately predict the movement of the currency.
Traders will make less mistakes and will be more confident in their trading.
So...
How was it done?A Support Vector Machine (SVM) algorithm, a machine learning technique is used as the approach to train the model with the input of sentiment analysis.
The collection of news headlines from different news sources is used to generate the sentiment value which is modelled together with the Forex movement through SVM in the creation of a real-world model for traders.
Example of how SVM works
Example of machine learning process
The Outcome
What is the Significance?
It was also found that the sentiment of the news will have great influence on the Forex movement
YES!!!
The accuracy is over 80% based on other studies
Traders in the future will benefit from this model as the world economy uncertainties would be more persistent in the coming years
I would like to express my deepest appreciation to my supervisor Dr. Low Yeh Ching who guided and supported me throughout this project.
