Online Program

Return to main conference page
Friday, October 19
Fri, Oct 19, 7:30 AM - 8:30 AM
Hall of Mirrors
Continental Breakfast and Speed Poster 2, Sponsored by Fifth Third Bank

K- Nearest Neighbor and Foreign Exchange Rate Forecasting (304959)

*Vindya Kumari Pathirana, University of Connecticut 

Keywords: Foreign Exchange Trading, k-Nearest Neighbor Algorithm, Mahalanobis Distance, ARIMA, Multi-step Ahead Time Series Forecasting

K- Nearest Neighbor Algorithms are among the most popular non-linear pattern recognition methods outperform that outperform the available linear forecasting methods for the high frequency foreign exchange data. In this work, we used five different foreign currencies, which are among the most traded currencies, to compare the performances of the k-NN algorithm with traditional Euclidean and Absolute distances to performances with the proposed Mahalanobis distance. The performances were compared in two ways: (i) forecast accuracy and (ii) transforming their forecasts in to a more effective technical trading rule. The results were obtained with real FX trading data, and the results showed that the method introduced in this work outperforms the other popular methods. Apart from the distance choice, we conducted a thorough investigation of optimal parameter choice with different distance measures. We also adopted the concept of distance based weighting to the NN and compared the performances with traditional unweighted NN algorithm based forecasting. The concept of dynamically adopted number of neighbors was also investigated.