An optimized SVM-k-NN currency exchange forecasting model for Indian currency market

This paper considers the prediction of currency exchange rate, volatility, and momentum prediction by exploring the capabilities of Support Vector Machine (SVM) and k-Nearest Neighbor (k-NN). In this work, the parameters such as penalty C and kernel gamma of SVM have been tuned with few optimization techniques such as random search, grid search, genetic algorithm, particle swarm optimization, ant colony optimization, firefly optimization, and BAT optimization algorithm. The final prediction has been obtained using k-NN by searching the neighborhood elements for either profit or loss. The performance of the proposed system has been tested with the Indian rupees with dollar (USD), British Pound (GBP), and Euro (EUR) for daily, weekly, and monthly in advance for prediction of currency exchange rate, volatility, and momentum in the currency market. The model BAT-SVM-k-NN has been found with the best forecasting ability based on performance measures such as mean absolute percentage error, root mean square error, mean squared forecast error, root mean squared forecast error, and mean absolute forecast error in comparison with other optimization techniques mentioned above.