Combining nonlinear independent component analysis and neural network for the prediction of Asian stock market indexes

Bandeira

2017

JRFM

Neural networks are well suited to predict future results of time series for various data types. This paper proposes a hybrid neural network model to describe the results of the database of the New York Stock Exchange (NYSE). This hybrid model brings together a self organizing map (SOM) with a multilayer perceptron with back propagation algorithm (MLP-BP). The SOM aims to segment the database into different clusters, where the differences between them are highlighted. The MLP-BP is used to construct a descriptive mathematical model that describes the relationship between the indicators and the closing value of each cluster. The model was developed from a database consisting of the NYSE Composite US 100 Index over the period of 2 April 2004 to 31 December 2015. As input variables for neural networks, ten technical financial indicators were used. The model results were fairly accurate, with a mean absolute percentage error varying between 0.16% and 0.38%.

Section: Implementation Of the Hybrid Modelmentioning

confidence: 99%

Section: Implementation Of the Hybrid Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Modeling NYSE Composite US 100 Index with a Hybrid SOM and MLP-BP Neural Model

Bandeira

2017

JRFM

“…Their experiment reveals that useful predictions can be made without using the extensive market data or knowledge. Dai, Wu and Lu (2012) proposed a time series prediction model by combining nonlinear independent component analysis (NLICA) and neural network to forecast Asian stock markets represented by Japanese and Chinese stock markets. Their results indicate that the proposed stock index prediction model can be a good alternative for Asian stock market indexes.…”

Section: Literature Reviewmentioning

confidence: 99%

Estimating the Turkish Sectoral Market Returns via Arbitrage Pricing Model under Neural Network Approach

Gökgöz¹,

Sezgin-Alp²

2014

IJEF

As an alternative to Capital Asset Pricing Model (CAPM), Arbitrage Pricing Theory (APT) is crucial in analyzing practical asset prices. APT provides a kind of multi factor market model which describes the expected returns with respect to macro-economic factors. In multifactor financial modeling, generally the traditional linear models are preferred. However, in the finance literature there are researches indicating the non-stationary and non-linearity of asset prices. For this purpose, in this paper the Artificial Neural Network (ANN) with Feed Forward Back Propagation algorithm has been employed to estimate the expected returns of the main sector indices of the İstanbul Stock Exchange, an emerging stock market, by using their relations with macroeconomic variables over 2003 and 2012 period. The forecasting ability of the ANN model is accessed using in sample and out of sample means square error (MSE) statistics and hypothesis test statistics testing whether there are differences between predicted returns and real returns. The results have revealed that the methodology based on ANN has a significant ability to estimate the different sectors in Turkish stock market with APT approach.

Proceedings of the 2015 Conference of the International Fuzzy Systems Association and the European Society for Fuzzy Logic and

“…Chang et al [3] developed an integrated system by combining dynamic time windows, case-based reasoning, and neural networks for stock trading decision support. In [4], a time-series prediction model was developed by combining non-linear independent component analysis (NLICA) and neural networks, which was proposed to forecast Asian stock markets. Bisoi and Dash [5] presented a simple feed-forward dynamic neural network comprising one or more layers of dynamic neurons for predicting stock price indices and profit from one day ahead to 30 days in advance.…”

Section: Introductionmentioning

confidence: 99%

Integration of mutual information and CRPSO-based fuzzy model for stock index forecasting

Yu¹,

Kim²

2015

In this paper, the integration of mutual information (MI) and fuzzy model is proposed to predict stock indexes with complex and non-linear characteristics. Technical indicators are considered as initial input candidates and significant inputs are determined by MI-based input selection method. To identify the structures and parameters of fuzzy models simultaneously, cooperative random learning particle swarm optimization (CRPSO), proposed by Zhao et al., is used. To confirm the effectiveness, the proposed method and comparison methods are applied to the Korea Composite Stock Price Index (KOSPI). The experimental results show that the proposed method, on average, outperforms other comparison methods.