Teaching–Learning Optimization Based Cascaded Low-Complexity Neural Network Model for Exchange Rates Forecasting

Abstract

An efficient hybrid forecasting model based on teaching–learning-based optimization cascaded with functional link artificial neural network (CFLANN-TLBO) is proposed in this paper. This hybrid method is mainly used for the prediction of the exchange of currency rates between one US Dollar (USD) to Indian Rupees (INR) and Canadian Dollar (CAD). In cascading FLANN model, computational complexity has reduced as well as the weights of the model optimized by TLBO algorithm to converge faster. The model’s performance is measured by determining the mean absolute percentage error (MAPE). The performance of the proposed model is also compared with other optimization techniques like cat swarm optimization (CSO), particle swarm optimization (PSO), and differential evolution (DE)-based cascaded FLANN. The proposed model performs better in comparison to cat swarm optimization (CSO), particle swarm optimization (PSO), and differential evolution (DE) with a higher accuracy.