Backwarding : An Overfitting Control for Genetic Programming in a Remote Sensing Application

Created by W.Langdon from gp-bibliography.bib Revision:1.4202

  author =       "Denis Robilliard and Cyril Fonlupt",
  title =        "Backwarding : An Overfitting Control for Genetic
                 Programming in a Remote Sensing Application",
  booktitle =    "Artificial Evolution 5th International Conference,
                 Evolution Artificielle, EA 2001",
  year =         "2001",
  editor =       "P. Collet and C. Fonlupt and J.-K. Hao and 
                 E. Lutton and M. Schoenauer",
  volume =       "2310",
  series =       "LNCS",
  pages =        "245--254",
  address =      "Creusot, France",
  month =        oct # " 29-31",
  publisher =    "Springer Verlag",
  ISBN =         "3-540-43544-1",
  DOI =          "doi:10.1007/3-540-46033-0_20",
  keywords =     "genetic algorithms, genetic programming",
  abstract =     "Overfitting the training data is a common problem in
                 supervised machine learning. When dealing with a remote
                 sensing inverse problem, the PAR, over-fitting prevents
                 GP evolved models to be successfully applied to real
                 data. We propose to use a classic method of over
                 fitting control by the way of a validation set. This
                 allows to go backward in the evolution process in order
                 to retrieve previous, not yet over fitted models.
                 Although this {"}backwarding{"} method performs well on
                 academic benchmarks, there is not enough improvement to
                 deal with the PAR. A new backwarding criterion is then
                 derived using real satellite data and the knowledge of
                 plausible physical bounds for the PAR coefficient in
                 the geographical area that is monitored. This leads to
                 satisfactory GP models and drastically improved
  notes =        "EA'01",

Genetic Programming entries for Denis Robilliard Cyril Fonlupt