Stock market prediction using hybrid multi-layer decomposition and optimized multi-kernel extreme learning machine

Abstract

The nancial time series data is a highly nonlinear signal and hence difficult to predict precisely. The prediction accuracy can be improved by linearizing the signal. In this paper, the nonlinear data sample is linearized by decomposing it into several Intrinsic Mode Functions (IMFs). A hybrid multi-layer decomposition technique is developed. The decomposition method proposed in this paper is composed of both Empirical Mode Decomposition (EMD) and Variational Mode Decomposition (VMD) methods individually. As a new contribution to the previous literature in this study, the VMD is used to decompose further the higher frequency signals obtained from the EMD-based decomposed signal. The result analysis shows that the double decomposition technique improves prediction accuracy. This is a new introduction to the eld of stock market prediction. The prediction accuracy of the proposed model is veri ed by applying it to three di erent stock market data. Historical data (closing price) is implemented to obtain one day ahead predicted closing price. Comparative analysis of other previously implemented methods like Back Propagation Neural Network (BPNN), Support Vector Machine (SVM), Arti cial Neural Network (ANN), and Extreme Learning Machine (ELM), along with the proposed method, is performed. Fire y algorithm is implemented for optimizing the kernel factors. It is observed that the proposed hybrid model outperformed other methods discussed in this study.