Name
Title | ||
|---|---|---|
An Arctan-Activated WASD Neural Network Approach to the Prediction of Dow Jones Industrial Average. |
Abstract | ||
|---|---|---|
Accurate prediction of the stock market index is a very challenging task due to the highly nonlinear characteristic of financial time series. For this reason, a single hidden-layer feed-forward neural network, activated by the arctan function, is proposed and investigated for predicting the Dow Jones Industrial Average. Then, a weights and structure determination (WASD) method is exploited to train the proposed arctan-activated neural network (termed arctan-activated WASD neural network). The relatively optimal weight and structure could be obtained by the presented WASD method. Numerical experiments are carried out based on huge amounts of historical data. The experimental results demonstrate the effectiveness and superior abilities of the arctan-activated WASD neural network for predicting the Dow Jones Industrial Average. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1007/978-3-319-59072-1_15 | ADVANCES IN NEURAL NETWORKS, PT I |
Keywords | Field | DocType |
Arctan,WASD neural network,Prediction,Dow Jones Industrial Average | Nonlinear system,Stock market index,Computer science,Artificial intelligence,Artificial neural network,Machine learning,Inverse trigonometric functions | Conference |
Volume | ISSN | Citations |
10261 | 0302-9743 | 0 |
PageRank | References | Authors |
0.34 | 4 | 5 |