Abstract | ||
---|---|---|
In this paper we evaluate the performance of backpropagation neural networks applied to the problem of predicting stock market prices. The neural networks are trained to approximate the mathematical function generating the semi-chaotic timeseries which represents the history of stock market prices in order to predict the values for the future. In contrast to previous investigations, the training data used in our experiments is not exclusively based on stock market prices, but also incorporates a variety of other economical factors. The prediction quality obtained is illustrated by presenting several simulation results. |
Year | DOI | Venue |
---|---|---|
1992 | 10.1007/BFb0024997 | IEA/AIE |
Keywords | Field | DocType |
stock market prediction,backpropagation networks,backpropagation,neural network | Training set,Time series,Function (mathematics),Computer science,Artificial intelligence,Artificial neural network,Backpropagation,Stock market prediction,Stock market,Machine learning | Conference |
ISBN | Citations | PageRank |
3-540-55601-X | 12 | 2.06 |
References | Authors | |
3 | 1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Bernd Freisleben | 1 | 12 | 2.06 |