Call for Paper - September 2020 Edition
IJCA solicits original research papers for the September 2020 Edition. Last date of manuscript submission is August 20, 2020. Read More

Time Series Regression Model for Prediction Of Closing Values of the Stock using an Adaptive NARX Neural Network

International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
Year of Publication: 2017
Saurabh Labde, Stuti Patel, Megh Shukla

Saurabh Labde, Stuti Patel and Megh Shukla. Time Series Regression Model for Prediction Of Closing Values of the Stock using an Adaptive NARX Neural Network. International Journal of Computer Applications 158(10):29-34, January 2017. BibTeX

	author = {Saurabh Labde and Stuti Patel and Megh Shukla},
	title = {Time Series Regression Model for Prediction Of Closing Values of the Stock using an Adaptive NARX Neural Network},
	journal = {International Journal of Computer Applications},
	issue_date = {January 2017},
	volume = {158},
	number = {10},
	month = {Jan},
	year = {2017},
	issn = {0975-8887},
	pages = {29-34},
	numpages = {6},
	url = {},
	doi = {10.5120/ijca2017912744},
	publisher = {Foundation of Computer Science (FCS), NY, USA},
	address = {New York, USA}


In the financial sector, the dependence on modern technologies has seen an increase since the last two decades. The advancements in the field of neural networks and machine learning has provided a number of financial tools. These tools often seen to form the basis of financial computations such as stock market prediction, bankruptcy prediction, risk assessment etc. In this paper, we propose a model based on an adaptive NARX neural network to predict the closing price of any stock. This is a non-linear auto regressive exogenous input model which uses delays in the input as well as the output acting as memory slots thereby increasing the accuracy of the prediction. This model uses a time series approach to analyze and predict the closing price . This NARX model is trained using three input values - the opening price of the stock , the highest price of the stock and the lowest price of the stock for the day . The target values are also fed to the network as it is a supervised learning model. Levenberg-marquardt algorithm has been used for training the network. The accuracy of the network is determined with the help of the mean squared error. In this model, we have made use of a closed loop with reduced input delays as well as a closed loop model for making predictions and the accuracy of each case is determined to analyze the working of a NARX neural network and to determine the optimum configuration.


  1. AbhishekKar (Y8021), Department of Computer Science and Engineering, IIT Kanpur, “Stock Prediction using Artificial Neural Networks”.
  2. Selvan Simon, ArunRaoot, International Journal on Soft Computing (IJSC), Vol.3 No.2, May 2012 ,“Accuracy Driven Artificial Neural Networks in Stock Market Prediction”.
  3. MarijanaZekic, University of JosipJurajStrossmayer in Osijek, “Neural Network Applications in Stock Market Predictions – A Methodolgy Analysis”.
  4. ParasharChandrashekharSoman, a thesis submitted to the Graduate School– New Brunswick Rutgers ,The State University of New Jersey, October 2008, “An Adaptive NARX Neural Network Approach for Financial Time Series Prediction”.
  5. “Application of artificial neural network for the prediction of stock market returns:The case of Japanese stock market”, Chaos,Solitons and Fractals, Volume 85,April 2016, Mingyue Qui, Yu Song, Fumio Akagi, Department of Systems Management, Fukuoka Institute of Technology.
  6. Images from google ,
  7. Bhagwant Chauhan, UmeshBidave, AjitGangathade, SachinKele, International Journal of Computer Science and Information Technologies, Vol 5(1),2014,904-907, “Stock Market Prediction Using Artificial Neural Networks”.
  8. Amin Hedayati Moghaddam, Moein Hedayati Moghaddam, “Journal of Economics, Finance and Administrative Science”, Science 21 (2016)89-93, “Stock market index prediction using artificial neural network”.
  9. AroshineMunasinghe,DajanaVlajic, Royal Institute of Technology DD143X, Bachelor’s Thesis in Computer Science, Thesis Supervisor: Pawel Herman, June 2015, “Stock market prediction using artificial neural networks”.
  10. “Neural Network Applications in Stock Market Predictions- A Methodology Analysis”, MarijanaZekic, MS, University of JosipJurajStrossmayer in Osijek.
  11. Xiaohua Wang, P.K.H. Phua, “Stock Market Prediction using neural networks: Does trading volume help in short-term prediction”, Neural Networks, 2003,Proceedings of the International Joint Conference on Neural Networks.
  12. AmirhosseinGhaznavi, Mohammad Aliyari, Mohammad Reza Mohammadi, International Research Journal of Applied and Basic Sciences,ISSN:2251-838X, Vol 10,2016,”Predicting Stock Price Changes of Tehran Artmis Company Using Radial Basis Function Neural Networks”.
  13. HakobGrigoryan. Bucharest University of Economic Studies, Buchrest, Romania, “Stock Market Prediction using Artificial Neural Networks, Case Study of TAL1T, Nasdaq OMX Baltic Stock.


Neural Networks, NARX ( Non-Linear Auto Regressive Exogenous inputs model), Training data, Target data, delays, mean - squared error.