Forecasting of Indian Rupee (INR) / US Dollar (USD) Currency Exchange Rate Using Artificial Neural Network

TL;DR

This paper demonstrates the successful application of artificial neural networks to forecast the INR/USD exchange rate, analyzing how input nodes, hidden nodes, and training sample size affect model accuracy.

Contribution

It investigates the impact of neural network parameters on exchange rate forecasting accuracy, providing insights into optimal model configurations.

Findings

Input nodes significantly influence performance.

Larger training samples reduce forecast errors.

Number of hidden nodes has less impact.

Abstract

A large part of the workforce, and growing every day, is originally from India. India one of the second largest populations in the world, they have a lot to offer in terms of jobs. The sheer number of IT workers makes them a formidable travelling force as well, easily picking up employment in English speaking countries. The beginning of the economic crises since 2008 September, many Indians have return homeland, and this has had a substantial impression on the Indian Rupee (INR) as liken to the US Dollar (USD). We are using numerational knowledge based techniques for forecasting has been proved highly successful in present time. The purpose of this paper is to examine the effects of several important neural network factors on model fitting and forecasting the behaviours. In this paper, Artificial Neural Network has successfully been used for exchange rate forecasting. This paper examines the effects of the number of inputs and hidden nodes and the size of the training sample on the in-sample and out-of-sample performance. The Indian Rupee (INR) / US Dollar (USD) is used for detailed examinations. The number of input nodes has a greater impact on performance than the number of hidden nodes, while a large number of observations do reduce forecast errors.