The stock market can be thought of as a highly
complex and adaptive system. Variation in share price of the
stock market can be considered as an indicator of the
economical trend of a country. Thus, forecasting the
behaviour of the stock market is at primary concern not only
of the business community but also of the policy markers of
a country. The fluctuation of share prices of the stock
market depend on external and internal factors. The major
difficulty in this regards is that most of these factors cannot
be quantified in a model. The results in this research clearly
illustrates that the historical data with external input factor
(such as the average exchange rate) increases the more
accurate share price forecasting ability of the neural
network model.
E.J.K.P Nandani : graduated from University of Ruhuna (2002-2008),
following B.Sc. (Special Mathematics) with specialization in Applied
Mathematics with a First Class Honours. and M.Phil. Degree in
(2009-2013) with title “Construction of a Back-propagation Artificial
Neural Network Model with Learning Rate Adaptation to Forecast the
Dynamical Behaviour of Colombo Stock Market" under the
supervision of Professor J.R. Wedagedara and Dr. M.K. Abeyratne,
Department of Mathematics, University of Ruhuna. At present I am a
PhD. Student, Wuhan Institute of Physics and Mathematics, Chinese
Academy of Sciences, China. In addition I am a Lecturer (on leave)
Department of Mathematics, University of Ruhuna, Sri Lanka.
Moreover, some research findings were the poster presented at the
5th and 9th Science Symposium of University of Ruhuna hold in 2009
and 2012 respectively. In 2015, I had a oral presentation in the 2nd
Ruhuna International science & Technology Conference title on
“Comparison of ARIMA and Neural Network Models for S&P SL 20
Index. My major teaching/research interests are on Neural Networks,
Mathematical Physics, Quantum Statistical Mechanics, Dynamical
systems, Control Theory, Time Series, Stochastic Differential
Equations and Computer Programming.
External Factor
Forecasting
Neural Network
S&P SL20 Index
The forecasting ability of the network model can be
improved had various socio-economic indicators (e.g.
exchange rate, export prices of commodity, cost of living
index, GDP etc.) been used as input factors in addition to
historical data.
[1] Suchira Chaigusin, Chaiyaporn Chirathamjaree, and
Judith Cclay- den. The use of neural networks in the
prediction of the stock exchange of thailand (set)
index. ECU Publications, pages 670–673, 2008.
[2] CHAN Man-Chung, WONG Chi-Ccccheong, and
LAM Chi Chung. Financial time series
forecasting by neural network using
conjugate gradient learning algorithm
andmultiple linear regression weight
initialization.
http://fmwww.bc.edu/cef00/papers/paper61.pdf, Jan
1988.
[3] Xuerong Mao. Stochastic Differential Equations and
their appli- cations. Chichester : Horwood, 1-89856-
326-8, 1997.
[4] J.R. Wedagedara and W.A.R. De Mel. Some politics
and statistical physics behind 1994 crash in colombo
stock market. pages 101-124, September 2006.
[5] http://ezinearticles.com/?expert=micheal-james.
[6] T.M.J.A.Cooray. Applied time series analysis and
forecasting. I Narosha, 2008.
[7] Roselina Sallehuddin, Siti Mariyam hj, Siti Zaiton
mohd, Hashim, and Ajith Abraham. Forecasting time
series data using hybridge grey ralational artificial
neural network and auto regressive inte- grated
moving average model. Neural Network World-Ics AS
CR2007, 6:573–625,.
[8] Holger R. Maier. and Graeme C. Dandy. Neural
networks for the prediction and forecasting of water
resources variables: a review of modelling issues and
applications. Environmental Modelling and Softwar,
15:101 – 124, 2000.
[9] Warren S. Sarle. Stopped training and other
remedies for overfitting. 27th Symposium on the
Interface, 1995.
[10] J.Dhar, Pankaj Agrawal, Viskas Singhal, Ajay Singh,
and Ranjan Kr.Murmu. Comparative study of
volatility forecasting between ann and hybrid models
for indian market. International Research Journal of
Finance and Economics, ISSN 1450-2887:68–
79,2010.
[11] E.J.K.P. Nandani, M.K. Mahinda, and J.R.
Wedagedara. Compar- ision of arima and neural
network models for sp sl(20) index. 2nd Ruhuna
INternational Science Technology Conference,
University of Ruhuna, SriLanka, Jan 22-23, 2015.
[12] Wei Huang, Shouyang Wang, Lean Yu, Yukun
Bao, and Lin Wang. A new computational method of
input selection for stock market forecasting with
neural networks. Springer-Verlag Berlin Heidelberg,
4:308–315, 2006.
[13] J.R. Wedagedara and E.J.K.P. Nandani. Designning a
decision as- sistant for bank loan classification using
back-propagation neural network. 5th Science
Symposium, University of Ruhuna,SriLanka, Feb 2009.
[14] E.J.K.P. Nandani and J.R. Wedagedara.
Comparative study of the forecasting ability of some
backpropagation artificial neuarl network models with
learning rate adaptation for colombo stock market.
Science Symposium , University of Ruhuna,
SriLanka, page 23, 2011.
[15] Wei Huang Shouyang Wang, Lean Yu, Yukun Bao, ,
and Ling Wan. A new computational method of input
selection for stock market forecasting with neural
networks. Neuro Computing, 4:308–315, 2006.
[16] Leandro S. maciel and Rosangela Ballini. Design a
neural network for time series financial forecasting:
Accuracy and robustness analysis. Institute de
Economia(IE), Universidade Estadual de
Campinas(UNICANP).
[17] Jingtao Yao, Hean lee Poh, and Teo Jasic. Foreign
exchange rates forecasting with neural networks. In
International Conference on Neural Information
Processing (Hong Kong), pages 754–759. Springer-
Verlag., 1996.
[18] Francesco Virili and Bernd Freisleben. Nonstationary
and data preprocessing for neural network predictions
of an economic time series. IEEEXplore-Conference
Publications on Neural Networks (IJCNN 2000),
5:129–134, 2000.
[19] Justin Werfel, Xiaohui Xie, and H.Sebastian Seung.
Learning curves for stochastic gradient descent in
linear feedforward networks. MIT Press Journal-
Neural Computation, 17:2699–2718, December 2005.
[20] Leon Bottou. Stochastic gradient learning in neural
networks.
http://leon.bottou.org/publications/pdf/nimes-
1991.pdf. [21] S. Rajasekaran and G.A. Vijayalakshmi Pai. Neural
Networks, Fuzzy Logic and Genetic Algorithm
(Synthesis and Application). Prentice- Hall of India
(pvt) Ltd,New Delhi, 2004.
[22] Anders Krogh and John A. Hertz. A simple weight
decay can improve generalization. In Advance in
Neural Information processing System, 4:950–957,
1995.
[23] http://tradertech.com.
[24] Iebeling Kaastra and Miltom Boydl. Design a model
for financila forecasting. Neuro Computing, 10:215–
236, 1996.
[25] Irina Klevecka and Janis Lelis. Pre-processing of input
data of neural networks: The case of forecasting
telecommunication network traffic. ISSN 0085-710,
2009.
[26] Steven Gonzalez. Neural networks for macroecomonic
forecasting: A complementary approach to linear
regression models.
http://publications.gc.ca/collections/collection2 008/f
in/F 21-8-2000-7E.pdf.
[27] B.H. Wang and P.M.Hui. The distribution and scaling
of fluctuations for hang seng index in hong kong stock
market. European Phsics Journal B,20:573–579, 2001.
