A Comparative Study of Time Series Prediction Based on Neural Network and the Ornstein-Uhlenbeck Process with Jumps

DOI

10.2991/ecae-17.2018.36How to use a DOI?

Keywords

stochastic time series; dynamic neural network; the ornstein-uhlenbeck process; stock prices

Abstract

The research of stock price prediction is very important. Traditionally, the stock price is usually processed as a time series. However, the modelling of such time series is extremely important and vital, and has been attracting the attention of both practitioners and researchers. In this paper, the dynamic neural network model and the OU process with jumps are used to analyze the stock prices respectively, and the two models are compared by fitting and prediction performance. The experimental results show that the OU process with jumps is superior to the dynamic neural network for stochastic time series prediction.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Yaohui Bai
AU  - Yan Tu
AU  - Huayang Li
PY  - 2017/12
DA  - 2017/12
TI  - A Comparative Study of Time Series Prediction Based on Neural Network and the Ornstein-Uhlenbeck Process with Jumps
BT  - Proceedings of the 2017 2nd International Conference on Electrical, Control and Automation Engineering (ECAE 2017)
PB  - Atlantis Press
SP  - 169
EP  - 173
SN  - 2352-5401
UR  - https://doi.org/10.2991/ecae-17.2018.36
DO  - 10.2991/ecae-17.2018.36
ID  - Bai2017/12
ER  -