Survival Analysis Using A Multi-Objective Evolutionary Algorithm

Created by W.Langdon from gp-bibliography.bib Revision:1.3872

@InProceedings{setzkorn:2005:CIMED,
  author =       "C. Setzkorn and A. F. Taktak and B. E. Damato",
  title =        "Survival Analysis Using A Multi-Objective Evolutionary
                 Algorithm",
  booktitle =    "Proceedings of the 2nd International Conference on
                 Computational Intelligence in Medicine and Healthcare -
                 CIMED",
  year =         "2005",
  keywords =     "genetic algorithms, genetic programming, evolutionary
                 algorithms, survival analysis",
  pages =        "224--230",
  address =      "Costa da Caparica, Lisbon, Portugal",
  month =        "29 " # jun # "-1 " # jul,
  URL =          "http://repository.liv.ac.uk/id/eprint/1194537",
  abstract =     "proposes a multi-objective evolutionary algorithm for
                 the extraction of radial basis function networks from
                 survival data. This type of artificial neural network
                 has a simpler structure than, for example, the
                 multi-layer perceptron, which has already been used for
                 survival analysis. The simpler structure of radial
                 basis function networks allows a faster model
                 extraction and better interpretation of the parameters
                 of the model. Multi-objective evolutionary algorithms
                 have several advantages over other optimisation methods
                 such as back-propagation. They can, for example, cope
                 better with feature interactions and noisy data.
                 Furthermore, they are capable of optimising several
                 objectives. This is important in the context of model
                 extraction, which is a multi-objective problem. It has
                 at least two objectives, which are the extraction of
                 (1) accurate and (2) simple models from data. Accurate
                 models are required to achieve good predictions. Model
                 simplicity is important to prevent overfitting, improve
                 the transparency of the models, and to save
                 computational resources. The proposed approach is
                 applied to two datasets. The extracted models achieve
                 good predictive performance.",
  notes =        "CIMED2005
                 http://www.uninova.pt/cimed2005/Programme%20Book.pdf",
}

Genetic Programming entries for Christian Setzkorn Azzam F G Taktak Bertil E Damato

Citations