Fuzzy ARTMAP neural network prediction of heart surgery mortality

Author(s): Carpenter, G.A. | Egbert, D. | Goodman, P. | Grossberg, S. | Grover, F. | Hammermeister, K. | Kaburlasos, V. | Marshall, G. | Reynolds, J.H. |

Year: 1992

Citation: Wang Institute Conference on Neural Networks, 48.

Abstract: A major national effort is underway to determine patterns of medical practice that most effectively result in favorable health outcomes. Databases arising from such effectiveness research may contain tens of thousands of cases and hundreds of variables intended to predict outcome status. Established statistical prediction algorithms may be suboptimal for such tasks because of obstacles arising fro massive number of cases, missing data, variable selection, multicollinearity, specification of important interactions, and sensitivity to erroneous values.

Topics: Machine Learning, Applications: Medical Diagnosis, Models: Fuzzy ARTMAP,

PDF download

Cross References