abhishek gupta et al, int.j.computer technology ... survey on stock market prediction using various...
TRANSCRIPT
A Survey on Stock Market Prediction Using Various
Algorithms
Abhishek Gupta,
Dr. Samidha D Sharma
Department of Information Technology, Department of Information Technology
NRI Institute of Information Science and Technology, Bhopal (M.P.) NRI Institute of Information Science and Technology, Bhopal (M.P.)
[email protected] [email protected]
Abstract
Stock market prediction is a technique of predicting the future value of the stock markets on the basis of the current and the
previous information available in the market.Stock market
prediction is an important issue for investigating in academic and
financial research. There are various techniques available for the prediction of the stock market value. Here in this paper a survey
of all the techniques and schemes for stock market prediction are
discussed and analysed.
Keywords: Stock Market, Prediction, Analysis.
Introduction
The prediction of stock market movement is an important
area of financial forecasting. Notwithstanding years of
study and the newest technology, it seems that no technique
has been exposed that constantly works. Essential analysis
frequently works best over longer periods of time, where
technical analysis is more appropriate for short term
trading. Researchers have made many attempts to predict
the performance of financial market .Many models or
approaches like artificial intelligence techniques such as
[1]Neural Network and Fuzzy Systems had proposed. However it is difficult to interpret their results .They are
unable to view the nature of interactions between technical
indicators and stock market fluctuations.
The difficulty with technical analysis is that a
complete pattern is required to make an accurate prediction
on the stock movement. Preferably, such a forecast should
be made before the pattern was completed to facilitate
prediction. For this task execute a Time-Delay Artificial
Neural Network known as Midas. Similar to Metis, Midas
receives input data including close, open, low and high
prices per day for a particular stock ticker over a period of
time, along with the equivalent trading quantity for every
day. This data is adequate for employing technical analysis
through Metis and is therefore used as the input data for
Midas as well [2].
The financial market unusual from a lot of substantial systems like we identify the weather is that the financial
market is an arrangement of complex feedback system.
What people expect prices to be concerns the prices they
examine and then the prices they examine then influences
how they are going to form their expectations about what
the prices will be in the after that stage. The market is
essentially an unresolved creature or an undecided
institution, it’s an institution where public trade risk,
exchange risk, and that’s why it’s present. And so if it were
probable to forecast it there would be no risk. In
individuals, I believe there cannot be any publicly
accessible system to predict a financial market.
Alternatively, neural networks have been originated useful
in stock price prediction [3-4]. Both feed forward and
recurrent neural networks have been investigated and good
results have been acquired. That means the calculation software would be very useful to support entity in
accomplishing a complete decision. In this paper, assuming
that it is feasible to predict markets, a prediction system is
extended using fuzzy neural networks with a learning
algorithm to predict the expectations stock assessments.
The system consists of numerous neural networks
components. These models are all used to discover the
relationships between unusual technical and economical
indices and the judgment to buy or sell stocks. The inputs
to the networks are practical and financial indices. The
amount produced of the system is the choice to buy and
sell. There are numerous neural network methods for stock
forecast, such as recurrent neural network, Time Series
method and Feed-forward neural network method, etc. [4].
When compared to these techniques, Fuzzy neural network
is a very valuable and efficient method to progression,
which is given in detailed in the later sections.
Stock market prediction is the act of trying to determine the
future value of a company stock or other financial
instrument traded on a financial exchange. The flourishing
calculation of a stock's future price could yield momentous
profit. Some consider that stock price movements are
governed by the random walk hypothesis and thus are
unpredictable. Others oppose and those with this
perspective possess a myriad of methods and technologies
Abhishek Gupta et al, Int.J.Computer Technology & Applications,Vol 5 (2),530-533
IJCTA | March-April 2014 Available [email protected]
530
ISSN:2229-6093
which purportedly allow them to gain future price
information [5].
Fuzzy Neural Networks
According to the mechanism of fuzzy logic control system,
the fuzzy neural network typically has 5 functional layers:
(1) Layer one is the input layer. (2) Layer two is the
fuzzification layer; (3) Layer three is the fuzzy reasoning
layer that may consist of AND layer and OR layer; (4) Layer four is the de-fuzzification layer; (5) Layer five is the
output layer. The design of a fuzzy neural network is
described in Figure.
Figure 1: Architecture of the fuzzy neural network
Usually, the fuzzy neural network maps crisp inputs xi
(I=1, 2… n) to crisp output yi (j=1, 2… m). A fuzzy neural network is constructed layer by layer according to fuzzy IF-
THEN rules, the fuzzy reasoning method, linguistic
variables, and the de-fuzzification scheme of a fuzzy reasoning method and the de-fuzzification scheme of a
fuzzy logic control system.
Each neuron in the fuzzification layer represents
an input membership function of the antecedent of a fuzzy
rule. One familiar method to execute this layer is to express
membership functions as discrete points. Thus for a fuzzy
rule "IF X1 is A1 and X2 is A2 ... THEN Y is B", A's
characterize the possibility distribution of the antecedent clause "X is A". Each of the hidden nodes is defined as a
fuzzy reference point in the input space. The function of
the de-fuzzification layer is for rule assessment. Each
neuron in this layer symbolizes a consequential proposition
"THEN Y is B" and its membership function can be
implemented by combining one or two sigmoid functions and linear functions [6].
Related Work
Nair, Mohandas and Sakthivel [7] proposes a
decision tree rough set hybrid system for stock market
prediction presents the design and performance evaluation of a hybrid decision tree- rough set based system for
predicting the next day’s trend in the Bombay Stock
Exchange (BSESENSEX). This system outperforms both
the neural network based system and the naïve bayes based
trend prediction system.
Lee, Anthony, Lin, Kao, and Chen [8] propose an
effective clustering approach to stock market prediction,
which combines the advantages of K-means and HAC, to
perform stock market prediction. This method consists of
three phases. First, it converts each financial report into a
feature vector and use HAC to divide them into clusters.
Second, for each cluster, it recursively apply K-means to
partition each cluster into sub-clusters so that most feature
vectors in each sub-cluster belong to the same class. Then,
for each sub-cluster, it chooses its centroid as the
representative feature vector. Finally, it employs the
representative feature vectors to predict the stock price movements.
Kim and Han [9] propose genetic algorithms
approach to feature discretization in artificial neural
networks for the prediction of stock price index. In this
Genetic Algorithm is employed not only to improve the
learning algorithm, but also to reduce the complexity in
feature space. GA optimizes simultaneously the connection
weights between layers and the thresholds for feature
discretization. The genetically evolved weights mitigate the
well-known limitations of the gradient descent algorithm.
This genetic based model outperforms the other
conventional models.
Kim [10] proposes artificial neural networks with
evolutionary instance selection for financial forecasting. He
uses a new hybrid model of ANN and genetic algorithms
(GAs) for instance selection. An evolutionary instance
selection algorithm reduces the dimensionality of data and may eliminate noisy and irrelevant instances. In addition, it
searches the connection weights between layers in ANN
through an evolutionary search. The genetically evolved
connection weights mitigate the well-known limitations of
gradient descent algorithm.GA-based learning and the
instance selection algorithm significantly outperforms the
conventional GA-based learning algorithm.
Matsui and Sato [11] propose neighborhood
evaluation in acquiring stock trading strategy using Genetic
algorithms. This method involves evaluation for
neighboring points of a genetic individual in fitness
Abhishek Gupta et al, Int.J.Computer Technology & Applications,Vol 5 (2),530-533
IJCTA | March-April 2014 Available [email protected]
531
ISSN:2229-6093
landscape as well as itself. It reduces the influence of the
singular points in the training phase and to improve the
profit in the testing phase.
Mahdi, Hamidreza and Homa [12] propose stock
market value prediction using neural networks. In this, two
kinds of neural networks, a feed forward multi layer
perception (MLP) and an Elman recurrent network, are
used to predict a company’s stock value based on its stock
share value history. The application of MLP neural network
is more promising in predicting stock value changes rather
than Elman recurrent network and linear regression
method. However, based on the standard measures Elman
recurrent network and linear regression can predict the
direction of the changes of the stock value better than the
MLP.
Shweta, Rekha and Vineet [13] propose predicting
future trends in stock market by decision tree rough-set
based hybrid system with HHMM. It presents a hybrid
system based on decision tree rough set, for predicting the
trends in the Bombay Stock Exchange (BSESENSEX) with the combination of Hierarchical Hidden Markov Model. It
also presents future trends on the bases of price earnings
and dividend. The data on accounting earnings when
averaged over many years help to predict the present value
of future dividends.
Sam Mahfoud and Ganesh Mani [14] propose
financial forecasting using genetic algorithms a new system
that utilizes genetic algorithms (GAs) to predict the future
performances of individual stocks. More generally, the
system extends GAs from their traditional domain of
optimization to inductive machine learning or
classification. The overall learning system incorporates a
GA, a niching method and several other components.
Chenoweth, Zoran and Lee [15] propose
embedding technical analysis into neural network based
trading systems. it compare an NN-based trading system
using a threshold-based pattern filtering technique to a system using a more sophisticated preprocessing technique
utilizing the ADX indicator to identify trends in the S&P
500 index. The results indicate that the ADX-based
directional filter used for preprocessing works better with
smaller ADX smoothing parameter values. However, the
simple threshold-based filtering technique still outperforms
the ADX based filtering technique with the rule-based
integration strategy.
Kuang and Jane [16] propose a hybrid model for
stock market forecasting and portfolio selection based on
ARX, grey system and RS theories. In this approach,
financial data are collected automatically every quarter and
are input to an ARX prediction model to forecast the future
trends of the collected data over the next quarter or half-
year period. The forecast data is then reduced using a
GM(1,N) model, clustered using a K-means clustering
algorithm and then supplied to a RS classification module which selects appropriate investment stocks by applying a
set of decision-making rules. Finally, a grey relational
analysis technique is employed to specify an appropriate
weighting of the selected stocks such that the portfolio’s
rate of return is maximized.
Chen, Leung and Daouk [17] propose application of neural networks to an emerging financial market:
forecasting and trading the Taiwan Stock Index. This
model predicts the direction of return on market index of
the Taiwan Stock Exchange, one of the fastest growing
financial exchanges in developing Asian countries.
Probabilistic neural network (PNN) is used to forecast the
direction of index return after it is trained by historical data.
Statistical performance of the PNN forecasts are measured
and compared with that of the generalized methods of
moments (GMM) with Kalman filter. Moreover, the
forecasts are applied to various index trading strategies, of
which the performances are compared with those generated
by the buy-and-hold strategy as well as the investment
strategies guided by forecasts estimated by the random
walk model and the parametric GMM models.
Connor, Niall and Madden [18] propose a neural
network approach to predicting stock exchange movements using external factors. In this substantial dataset has been
compiled with daily values of the DJIA, derived technical
indicators, and external indicators and profit-based
evaluation procedure for financial prediction systems is
proposed, based on simulations with a simple trading
strategy, it is more meaningful than evaluating systems
based on the error between predicted and actual values, as
is sometimes done. A neural network approach is shown to
be successful in predicting movements of the DJIA,
provided that external factors are considered. These results
may be used as a baseline against which to compare other
prediction techniques in the future.
Yang, Min and Lin [19] propose the application of
fuzzy neural networks in stock price forecasting based on
Genetic Algorithm discovering fuzzy rules. This paper
provides a method to improve black box model considering
problems existed in its application. The improvement is achieved mainly by applying GA (Genetic Algorithm) in
fuzzy systems to discover rules, eliminate errors or invalid
rules caused by noisy data, and thus form valid sets of
rules. Evaluation of the rule sets, as that of the whole
prediction model, is performed through known knowledge
and theories. At last, fuzzy reasoning approach is used
based on the rule sets to predict price trend of stock market.
Conclusion
Although there are various techniques implemented for the
prediction of stock market. Here in this paper a complete
survey of all the stock market prediction techniques are
given. On the basis of existing technique used for the future
prediction of stock market a new technique for the
prediction is proposed which provides close prediction of
stock market.
Abhishek Gupta et al, Int.J.Computer Technology & Applications,Vol 5 (2),530-533
IJCTA | March-April 2014 Available [email protected]
532
ISSN:2229-6093
References:
[1] Witold Pedrycz, Abraham Kandel and Yan-Qing Zhang, “Neuro-
fuzzy Systems,” In Fuzzy Systems: Modeling and Control, pp. 312–
363, 1999.
[2] Szu Han Chang and Michael Mallon. “Predicting the Stock Market
through Preprocessing Historical Data via an Expert System as Input
to a Time-Delay Artificial Neural Network”, Master Thesis Boston
University Department of Computer,2008
[3] Lee, C.H and Park, K.C. “Prediction of monthly transition of the
composition stock price index using recurrent back-propagation”,
Int... Conf. On Artificial Neural Networks, Brighton, UK, pp.1629-
1632. 1992
[4] D.T.Pham and X.Liu, “Neural Networks for Identification,
Prediction and Control”, Springer, 1999.
[5] Fischer Black and Myron Scholes, “The Pricing of Options and
Corporate Liabilities" The Journal of Political Economy, 81(3), pp.
637-654, 1973.
[6] Y.-Q. Zhang and A. Kandel, “Compensatory Genetic Fuzzy Neural
Networks and Their Applications,” Series in Machine Perception
Artificial Intelligence, World Scientific, 30, 1998.
[7] Nair, Binoy B., V. P. Mohandas, and N. R. Sakthivel. "A decision
tree-rough set hybrid system for stock market trend prediction."
International Journal of Computer Applications 6.9 (2010): 1-6.
[8] Lee, Anthony JT, Ming-Chih Lin, Rung-Tai Kao, and Kuo-Tay
Chen. "An Effective Clustering Approach to Stock Market
Prediction." In PACIS, p. 54. 2010.
[9] Kim, K. J., & Han, I.”Genetic algorithms approach to feature
discretization in artificial neural networks for the prediction of stock
price index”. Expert systems with applications, 19(2), 125-132.
2000.
[10] Kim, K. J,”Artificial neural networks with evolutionary instance
selection for financial forecasting”. Expert Systems with
Applications, 30(3), 519-526. . 2006
[11] Matsui, Kazuhiro, and Haruo Sato. "Neighborhood evaluation in
acquiring stock trading strategy using genetic algorithms." Soft
Computing and Pattern Recognition (SoCPaR), 2010 International
Conference of. IEEE, 2010.
[12] Mahdi Pakdaman Naeini, Hamidreza Taremian and Homa Baradaran
Hashemi. “Stock Market Value Prediction Using Neural Networks”,
International Conference on Computer Information Systems and
Industrial Management Applications (CISIM), pp. 132-136, 2010
[13] Shweta Tiwari, Rekha Pandit and Vineet Richhariya, “Predicting
future trends in stock market by decision tree rough-set based hybrid
system with HHMM”, International Journal of Electronics and
Computer Science Engineering, ISSN- 2277-1956, 1( 3) , pp. 1578 –
1587, 2012.
[14] Sam Mahfoud and Ganesh Mani, “Financial Forecasting using
Genetic Algorithms”, Applied Artificial Intelligence, 10: pp. 543-
565, 1996.
[15] Chenoweth, Tim, Zoran Obradovic, and Sauchi Stephen Lee.
"Embedding technical analysis into neural network based trading
systems." Applied Artificial Intelligence 10.6 pp.523-542.1996
[16] Huang, Kuang Yu, and Chuen-Jiuan Jane. "A hybrid model for stock
market forecasting and portfolio selection based on ARX, grey
system and RS theories." Expert Systems with Applications
36(3),pp.5387-5392.2009
[17] Chen, An-Sing, Mark T. Leung, and Hazem Daouk. "Application of
neural networks to an emerging financial market: forecasting and
trading the Taiwan Stock Index." Computers & Operations Research
30(6), pp. 901-923.2003
[18] O’Connor, Niall, and Michael G. Madden. "A neural network
approach to predicting stock exchange movements using external
factors." Knowledge-Based Systems 19(5) pp.371-378.2006
[19] Yang, Kongyu, Min Wu, and Jihui Lin. "The application of fuzzy
neural networks in stock price forecasting based On Genetic
Algorithm discovering fuzzy rules." In Natural Computation (ICNC), 2012 Eighth International Conference on, pp. 470-474.
IEEE, 2012.
Abhishek Gupta et al, Int.J.Computer Technology & Applications,Vol 5 (2),530-533
IJCTA | March-April 2014 Available [email protected]
533
ISSN:2229-6093