DOI: 10.4018/IJAEC.2016010102
Copyright © 2016, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
International Journal of Applied Evolutionary Computation
Volume 7 • Issue 1 • January-March 2016
Prediction of Financial Time Series
Data using Hybrid Evolutionary
Legendre Neural Network:
Evolutionary LENN
Rajashree Dash, Siksha ‘O’ Anusandhan University, Bhubaneswar, India
Pradipta Kishore Dash, Siksha ‘O’ Anusandhan University, Bhubaneswar, India
ABSTRACT
In this paper a predictor model using Legendre Neural Network is proposed for one day ahead
prediction of financial time series data. The Legendre Neural Network (LENN) is a single layer
structure that possess faster convergence rate and reduced computational complexity by increasing
the dimensionality of the input pattern with a set of linearly independent nonlinear functions. The
parameters of the LENN model are estimated using a Moderate Random Search Particle Swarm
Optimization Method (HMRPSO). The HMRPSO is a variant of PSO that uses a moderate random
search method to enhance the global search ability of particles and increases their convergence rates by
focusing on valuable search space regions. Training LENN using HMRPSO has also been compared
with Particle Swarm Optimization (PSO) and Differential Evolution (DE) based learning of LENN
for predicting the Bombay Stock Exchange and S&P 500 data sets.
KEywoRDS
BP Learning, Differential Evolution, Legendre NN, Moderate Random Search PSO, Particle Swarm Optimization
1. INTRoDUCTIoN
The increasing globalization of financial market attracts a lot of investors for investing stocks in stock
market. Suitable investment will help in getting more profit with less risk. But stock prices being
time series are noisy, random and volatile in nature. Predicting such highly fluctuating and irregular
stock prices is usually subject to large errors. The traditional statistical models used for analysis of
such time series data were simple, but suffered from several shortcomings due to the nonlinearity of
data. Hence developing more realistic models to predict stock price more effectively and accurately
is a great interest of research in financial data mining.
In this paper, a Legendre Neural Network (LENN) is proposed for one day ahead prediction
of financial time series data. Further the parameters of the network are estimated using a hybrid
Moderate Random Search Particle Swarm Optimization Method (HMRPSO). The Hybrid MRPSO
(HMRPSO) uses a moderate random search method to enhance the global search ability of particles
and increases their convergence rates by focusing on valuable search space regions. It seeks a balance
between the exploration of new regions and the exploitation of the already sampled regions in the
solution spaces. The proposed learning scheme has been also compared with other learning schemes
such as Backpropagation (BP), Particle Swarm Optimization (PSO) and Differential Evolution (DE)
algorithm. To test the model performance, two well-known stock market indices namely: Bombay
16