Open AccessForecasting Foreign Currency Exchange Price using Long Short-Term Memory with K-Nearest Neighbor Method
Author(s) -
Rudra Kalyan Nayak,
S.Y.H. Pavitra,
Rajesh Kumar Tripathy,
K. Prathyusha
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.b3551.129219
Subject(s) - currency , exchange rate , mean absolute percentage error , foreign exchange market , foreign exchange , econometrics , term (time) , mean squared error , database transaction , long memory , economics , computer science , statistics , monetary economics , mathematics , physics , quantum mechanics , programming language , volatility (finance)
With the growing population in the world, economic stability varies day by day. In case of India all banking transaction rules and regulations are taken by Reserve bank of India (RBI) whereas for other countries it is different. Therefore numerous academicians have projected their research on forecasting the currency exchange rate for diverse countryside. Foreign currency exchange rate prediction is a very pivotal task for international market. Hence researchers have explored different methods for predicting foreign currency exchange rate. In this work, we have taken Indian rupees (INR) with two different country’s data set such as Japanese yen (JPY) andChinese Yuan (CNY)for daily, weekly and monthlyprediction beforehand. We implemented a hybrid model oflong short-term memory (LSTM) with K-nearest neighbour (KNN) which gives better opening price prediction accuracy on our dataset. The accuracy of the prediction results are measured by the help of performance standards such as mean absolute percentage error (MAPE) and root mean square error (RMSE).