Chaos, Solitons and Fractals 85 (2016) 1–7 Contents lists available at ScienceDirect Chaos, Solitons and Fractals Nonlinear Science, and Nonequilibrium and Complex Phenomena journal homepage: www.elsevier.com/locate/chaos Application of artiﬁcial neural network for the prediction of stock market returns: The case of the Japanese stock market Mingyue Qiu * , Yu Song, Fumio Akagi Department of Systems Management, Fukuoka Institute of Technology, 3-30-1 Wajiro-higashi, Higashi-ku, Fukuoka 811-0295 Japan article info Article history: Received 7 December 2015 Accepted 6 January 2016 Keywords: Finance Artiﬁcial intelligence Artiﬁcial neural network Genetic algorithm Simulated annealing Japanese stock market abstract Accurate prediction of stock market returns is a very challenging task because of the highly nonlinear nature of the ﬁnancial time series. In this study, we apply an artiﬁcial neural network (ANN) that can map any nonlinear function without a prior assumption to predict the return of the Japanese Nikkei 225 index. (1) To improve the effectiveness of prediction algorithms, we propose a new set of input variables for ANN models. (2) To verify the pre- diction ability of the selected input variables, we predict returns for the Nikkei 225 index using the classical back propagation (BP) learning algorithm. (3) Global search techniques, i.e., a genetic algorithm (GA) and simulated annealing (SA), are employed to improve the prediction accuracy of the ANN and overcome the local convergence problem of the BP algorithm. It is observed through empirical experiments that the selected input vari- ables were effective to predict stock market returns. A hybrid approach based on GA and SA improve prediction accuracy signiﬁcantly and outperform the traditional BP training algorithm. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction To revive the Japanese economy, the Japanese govern- ment has recently developed many signiﬁcant economic strategies, and each strategy is closely related to the Japanese stock market. As the most widely used market index for the Tokyo Stock Exchange, the Nikkei 225 in- dex, also known as the Nikkei average or simply Nikkei, is a benchmark that is used to evaluate the Japanese econ- omy. Forecasting the stock return of the Nikkei 225 index is an important ﬁnancial subject that has attracted sig- niﬁcant attention in major ﬁnancial markets around the world. The purpose of this paper is to apply an artiﬁcial neural network (ANN) to forecast the return of the Nikkei 225 index. * Corresponding author. Tel.: +81 80 9062 7301; fax: +81 92 606 0756. E-mail addresses: qmy1175116@126.com, p15x1001@bene.ﬁt.ac.jp (M. Qiu), song@ﬁt.ac.jp (Y. Song), akagi@ﬁt.ac.jp (F. Akagi). It has been widely accepted by many studies that non- linearity exists in ﬁnancial markets and that an ANN can be used effectively to uncover this relationship [1]. McCul- loch and Pitts [2] created a computational model for neu- ral networks based on mathematics and algorithms, and the application of ANNs to ﬁnancial and investment deci- sions has been examined by researchers for many years. Compared to regression or the passive buy-and-hold strat- egy, Motiwalla and Wahab [3] found that ANN models are more successful in predicting returns. Enke and Thaworn- wong [1] used neural network models for level estimation and classiﬁcation. They showed that the trading strategies guided by a neural network classiﬁcation model can gener- ate higher proﬁts than any other model. Hodnett and Hsieh [4] utilized two ANN learning rules to forecast the cross- section of global equity returns. Their ﬁndings support the use of ANNs for ﬁnancial forecasting. Application of ANNs has become the most popular machine learning method, and it has been proven that such an approach can outper- form conventional methods [5-13]. http://dx.doi.org/10.1016/j.chaos.2016.01.004 0960-0779/© 2016 Elsevier Ltd. All rights reserved.