A B-P ANN Commodity Trader

Part of Advances in Neural Information Processing Systems 3 (NIPS 1990)

Authors

Joseph E. Collard

Abstract

An Artificial Neural Network recognize a buy/sell particular commodity Propagation of errors algorithm was used to encode the the Long/Short desired output and 18 fundamental variables plus 6 (or 18) Trained on one technical variables into year of past data to predict long/short market positions in the future that would have made $10,301 profit on an investment of less than$1000. relationship between