Build Prediction Models for Gold Prices Based on Back-Propagation Neural Network

DOI

10.2991/msam-15.2015.35How to use a DOI?

Keywords

gold price; back propagation neural network (BPN); Principal Component Regression (PCR); Multiple Regression (MR); technical index

Abstract

In recent years, international gold prices have been constantly rising, gold investment and preserve (or even appreciation) effects have been widely concerned by the market. Whether it is based on speculation, investment or hedging purposes, the gold has been incorporated into the asset allocation by many investors, which has become another important investment in addition to foreign currency, funds, stocks and securities. Therefore, this paper discusses how to construct a prediction model for gold prices to understand the future gold price trend, and to provide a reference for experts and investors. Firstly, we collect historical data of gold prices from web database, and draw a tendency chart to observe the trend of gold prices; then we use technical index formula of share price to calculate the five technical index values of gold as an independent variable and the price of gold the next day as a dependent variable, and build three prediction models including back-propagation neural network (BPN), Principal Component Regression (PCR) and Multiple Regression (MR). The study indicates that BPN’s predictive ability is better than other models.

Copyright

© 2015, 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  - Chingpei Lin
PY  - 2015/08
DA  - 2015/08
TI  - Build Prediction Models for Gold Prices Based on Back-Propagation Neural Network
BT  - Proceedings of the 2015 International Conference on Modeling, Simulation and Applied Mathematics
PB  - Atlantis Press
SP  - 155
EP  - 158
SN  - 1951-6851
UR  - https://doi.org/10.2991/msam-15.2015.35
DO  - 10.2991/msam-15.2015.35
ID  - Lin2015/08
ER  -