Neural network ensemble strategies for financial decision applications Export

@article{2005_WestDellanaQian, abstract = {Considerable research effort has been expended to identify more accurate models for decision support systems in financial decision domains including credit scoring and bankruptcy prediction. The focus of this earlier work has been to identify the single best prediction model from a collection that includes simple parametric models, nonparametric models that directly estimate data densities, and nonlinear pattern recognition models such as neural networks. Recent theories suggest this work may be misguided in that ensembles of predictors provide more accurate generalization than the reliance on a single model. This paper investigates three recent ensemble strategies: crossvalidation, bagging, and boosting. We employ the multilayer perceptron neural network as a base classifier. The generalization ability of the neural network ensemble is found to be superior to the single best model for three real world financial decision applications. (c) 2004 Elsevier Ltd. All rights reserved.}, author = {West, D. and Dellana, S. and Qian, J. X.}, citeulike-article-id = {652073}, journal = {Computers and Operations Research}, keywords = {bagging, bankruptcy, boosting, bootstrap, classification, credit, distress, ensembles, models, neural-networks, prediction, scoring}, number = {10}, pages = {2543--2559}, posted-at = {2006-05-18 18:31:12}, priority = {0}, title = {Neural network ensemble strategies for financial decision applications}, volume = {32}, year = {2005} }

TY - JOUR ID - 2005_WestDellanaQian L3 - citeulike-article-id:652073 N2 - Considerable research effort has been expended to identify more accurate models for decision support systems in financial decision domains including credit scoring and bankruptcy prediction. The focus of this earlier work has been to identify the single best prediction model from a collection that includes simple parametric models, nonparametric models that directly estimate data densities, and nonlinear pattern recognition models such as neural networks. Recent theories suggest this work may be misguided in that ensembles of predictors provide more accurate generalization than the reliance on a single model. This paper investigates three recent ensemble strategies: crossvalidation, bagging, and boosting. We employ the multilayer perceptron neural network as a base classifier. The generalization ability of the neural network ensemble is found to be superior to the single best model for three real world financial decision applications. (c) 2004 Elsevier Ltd. All rights reserved. IS - 10 JF - Computers and Operations Research EP - 2559 TI - Neural network ensemble strategies for financial decision applications VL - 32 SP - 2543 KW - bagging KW - bankruptcy KW - boosting KW - bootstrap KW - classification KW - credit KW - distress KW - ensembles KW - models KW - neural-networks KW - prediction KW - scoring AU - West, D AU - Dellana, S AU - Qian, JX PY - 2005/// ER -