Nature inspired soft computing techniques for prediction of indian currency exchange market

Abstract

The economy of a country is the strength of a country. For the economic growth of a newlinecountry global trading is very much required. Each country has its own currency value to newlinetrade with other country, thereby the currency exchange is essential. The value of the newlinecurrency varies from country to country therefore there is an organisation Foreign newlineExchange (Forex) which finds out the exchange rate of each country s currency. The value newlineof currency exchange price of any two countries is known as exchange rate. Exchange newlineprice of each two country s varies every hour and the value of a currency depends on newlineGross Domestic Product (GDP), inflation, government economic and trade policies, newlinegeopolitics, trade war and many more factors like social, political and foreign policies of a newlinecountry. Therefore, exchange rate of any two countries are volatile, non-static and newlinenonlinear. For the international business, prediction of the exchange price is very much newlinehelpful but it s very difficult as the exchange rate is volatile. Economists are using many newlinestatistical methods to predict the exchange rate. The statistical methods uses historical newlinedataset to understand the pattern of the price value changes. By using the current statistical newlinemodel to predict the price exchange rates the usual nature of these concerned data cannot newlinebe used adequately. For predicting the better accuracy now a day s Machine Learning newline(ML) techniques are used to understand the hidden relationship between the every day s newlinedata in time series dataset. The aim of this research is to find a good ML technique for newlinemore accuracy in less time than the previously used techniques. newlineThe basic objective of this research work is to find out a prediction model which newlinecan predict the future open price by taking input of current day s details. The models are newlineexperimented for predicting price in short range and long range in advance. The short newlinerange is one day, three days, one week, and one fortnight in advance whereas, long range newlineis one month, forty-five days, two months and one year in