![]() ![]() In: 9th International Conference on Neural Information Processing, vol. 1, pp. Jordanov, I.: Neural Network Training and Stochastic Global Optimization. Springer, Heidelberg (2004)Ĭaldwell, R.B.: Performances Matrics for Neural Network-based Trdaing System Development. In: Boulicaut, J.-F., Esposito, F., Giannotti, F., Pedreschi, D. KeywordsĪkbani, R., Kwek, S., Japkowwicz, N.: Appying Support Vector Machines to Imbalanced Datasets. The algorithms with the modified error functions introduced by this study produced better predictions. The algorithms developed in this study were employed to predict the trading signals of day ( t+1) of the Australian All Ordinary Index. A global optimization algorithm was employed to train these networks. An adjustment relating to the contribution from the historical data used for training the networks, and the penalization of incorrectly classified trading signals were accounted for when modifying the OLS function. In this paper, new algorithms were developed based on the structure of feedforward neural networks and a modified Ordinary Least Squares (OLS) error function. Most commonly used classification techniques are not suitable to predict trading signals when the distribution of the actual trading signals, among theses three classes, is imbalanced.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |