Financial Time Series Forecasting using Back Propagation Neural Network and Deep Learning Architecture
Richa Handa1, A.K. Shrivas2, H.S. Hota3
1Richa Handa, Department of Computer Science, Dr. C.V. Raman University, Bilaspur (C.G.), India.
2A.K. Shrivas, Department of Computer Science, Dr. C.V. Raman University, Bilaspur (C.G.), India.
3H. S. Hota, Department of Computer Science and Application, Atal Bihari Vajpayee University, Bilaspur (C.G.), India

Manuscript received on 01 April 2019 | Revised Manuscript received on 06 May 2019 | Manuscript published on 30 May 2019 | PP: 3487-3492 | Volume-8 Issue-1, May 2019 | Retrieval Number: A2137058119/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Artificial neural network is widely used for time series data which have got impact in today’s economy with advancement of new artificial techniques. In this study we have used three time series data i.e.BSE30 stock data, INR/USD Foreign Exchange (FX) data and Crude Oil Data for prediction. Many linear and non linear models have been developed for these time series prediction. Our approach in this work is to comprise a robust predictive model for next day ahead prediction. In this paper we have done comparative study of traditional artificial neural network: Error Back Propagation Network (EBPN) and Deep Neural Network: Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) for financial time series prediction. It has been observed that Deep Neural Network is outperforming the traditional Neural Network EBPN. The performances of models are measured by error measures: Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE).
Index Terms: Deep Neural Network, Error Back Propagation Network (EBPN), Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN).

Scope of the Article: Deep Learning