AN ARTIFICIAL INTELLIGENCE -BASED FOREX TRADING PREDICTION SOLUTION FOR KENYA
A INDIVIDUAL CAPSTONE PROJECT FOR MASTERS IN DATA SCIENCE AND ARTIFICIAL INTELLIGENCE
Keywords:
Data ScienceAbstract
In the realm of forex trading, accurate currency price predictions are paramount for
successful trading strategies. However, the challenge of inaccurate predictions persists,
particularly in regions like Kenya. This academic report delves into the extent of this
problem in the Kenyan context, exploring the domain of time series forecasting and
predictive analytics. The proposed solution employs supervised learning, chosen for its
ability to learn patterns from labelled historical data. The report outlines the specific
machine learning algorithm employed, focusing on developing a predictive model for
currency price movements using historical forex data and economic indicators.
Challenges in data acquisition are discussed, alongside the steps involved in data
access, incorporation of economic indicators, and dataset requirements. Detailed
explanations of data processing techniques, including time series analysis, feature
engineering, and data normalization, are provided, along with insights into the use of
time series forecasting models and classification models, such as LSTM, and their
suitability for the task. Deployment considerations, including the deployment platform
and how traders can utilize the model to inform their trading strategies, are explored,
along with challenges and limitations in real-world scenarios.
References
Brownlees, C., & Gallo, G. M. (2006). Financial econometric analysis at ultra-high
frequency: Data handling concerns. Journal of Financial Econometrics, 4(2), 339-
[Online]. Available: https://doi.org/10.1093/jjfinec/nbl012
National Bank of Kenya. (n.d.). Economic indicators. [Online]. Available:
https://www.nationalbank.co.ke/
Ngugi, R. W., & Muraya, P. (2018). Economic factors influencing exchange rate
volatility in Kenya. Journal of Business and Economic Development, 3(1), 7-19.
[Online]. Available: https://doi.org/10.15640/jbed.v3n1a2
OpenAI. (n.d.). OpenAI API. [Online]. Available: https://openai.com/api
Python Software Foundation. (2021). Python Language Reference, version 3.9.
[Online]. Available: https://docs.python.org/3.9/reference/index.html
Lipton, A., & Stein, M. (2018). Deep learning for time series forecasting. Packt
Publishing Ltd.
Prado, M. L., & Kritzman, M. P. (2019). Introduction to the mathematics of
financial derivatives. Cambridge University Press.
Tsay, R. S. (2013). Analysis of financial time series. John Wiley & Sons.
Yiu, M. S., & Wong, C. Y. (2017). AI techniques for financial time series
prediction. IGI Global.
Zumbach, G. O. (2012). Forex trading using intermarket analysis . John Wiley
& Sons.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural
computation, 9(8), 1735-1780.
Le, H., Nguyen, G., & Nguyen, H. (2019). A deep learning approach for stock
price forecasting. Journal of Economics and Development, 21(2), 66-79.
Saltoğlu, B., & Oğuz, F. (2017). Forecasting stock prices of Turkish companies
using Google trends data and machine learning techniques. International Journal
of Economics and Financial Issues, 7(2), 583-595.
Wu, C., Xu, Y., & Li, S. (2019). A hybrid model for stock price prediction using
LSTM and ARIMA. Symmetry, 11(5), 676.
Zhang, Y., & Wu, Q. (2020). Financial time series forecasting using deep
learning and extreme gradient boosting. Journal of Forecasting, 39(7), 1093-
Kaggle. (n.d.). Datasets - Forex. [Online]. Available:
https://www.kaggle.com/datasets?tags=forex
Kenya National Bureau of Statistics. (n.d.). Economic surveys. [Online].
Available: https://www.knbs.or.ke/economic-surveys/
TensorFlow. (n.d.). TensorFlow documentation. [Online]. Available:
https://www.tensorflow.org/guide
Towards Data Science. (n.d.). Articles on machine learning and AI. [Online].
Available: https://towardsdatascience.com/
World Bank. (n.d.). Data - Kenya. [Online]. Available: