Title: Efficient forecasting of financial time-series data with virtual adaptive neuro-fuzzy inference system

Authors: Sarat Chandra Nayak; Bijan Bihari Misra; Himansu Sekhar Behera

Addresses: Department of Computer Science & Engineering, Silicon Institute of Technology, Bhubaneswar 751024, Odisha, India ' Department of Information Technology, Silicon Institute of Technology, Bhubaneswar 751024, Odisha, India ' Department of Computer Science Engineering & Information Technology, Veer Surendra Sai University of Technology, Burla 768018, Odisha, India

Abstract: Uncertainties and non-linearity associated with the stock index make it difficult to predict its behaviour and hence it remains a challenging task for researchers. Newly developed intelligent machine learning techniques have been applied to this area and these have established as efficient forecasting models. This paper presents a Virtual Adaptive Neuro-Fuzzy Inference System (VANFIS) for efficient forecasting of stock market indices. VANFIS works in a virtual environment where the Adaptive Neuro-Fuzzy Inference System (ANFIS) is exposed to virtual data positions to infer the future stock price. This model does not take any actual data at any point of time as its input, but works completely in the virtual environment. To validate the performance of the proposed model, 15 years' data from ten stock markets are taken and five different performance metrics are evaluated. Simulation results show that VANFIS significantly improves forecasting performance in comparison to ANFIS.

Keywords: stock market forecasting; financial time series; virtual ANFIS; adaptive neuro-fuzzy inference systems; VDP; virtual data position; linear extrapolation; neural networks; fuzzy logic; stock prices; simulation.

DOI: 10.1504/IJBFMI.2016.080132

International Journal of Business Forecasting and Marketing Intelligence, 2016 Vol.2 No.4, pp.379 - 402

Received: 27 Jun 2016
Accepted: 29 Jul 2016
Published online: 03 Nov 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article