- author = "S. Openshaw and I. Turton",
- title = "Building New Spatial Interaction Models Using Genetic Programming",
- booktitle = "Evolutionary Computing, AISB workshop",
- year = "1994",
- editor = "T. C. Fogarty",
- address = "Leeds, UK",
- month = "11-13 " # apr,
- organisation = "AISB",
- note = "unpublished",
- keywords = "genetic algorithms, genetic programming, FORTRAN77, Cray-YMP",
-
URL = "
http://www.geog.leeds.ac.uk/papers/94-1/94-1.pdf",
-
URL = "
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.140.6775",
-
URL = "https://link.springer.com/content/pdf/bfm%3A978-3-540-48999-3%2F1.pdf",
-
DOI = "
doi:10.1007/3-540-58483-8",
- size = "10 pages",
- abstract = "Building computer models of human geographic systems is difficult because of the poverty of applicable theory that can be used to specify well formed and soundly based mathematical models. As a result there has been little progress in this area since the late 1960's when mathematical model building procedures based on entropy-maximising methods were first applied to spatial interaction data. In the mid-1980's, attention was focused on the use of genetic algorithms to explore the universe of different models that could be built from the available symbolic pieces: that is there is a very very large number of permutations of the model building blocks: viz. a mix of unary and binary operators, unknown parameters and observed variables, and rules of syntax for well formed equations. The so-called Automated Modeling System (AMS) worked but never fulfilled its expectations. The paper revisits the problem but re-casts the AMS approach into a genetic programming framework. Some preliminary results based on Cray-YMP runs are reported.",
-
notes = "Not included in Springer's LNCS Volume 865, 1994,
DOI:10.1007/3-540-58483-8. However the editor confirms
the work was presented at the workshop and this is
confirmed by page vii of the preface
Proceedings of the Workshop on Artificial Intelligence and Simulation of Behaviour Workshop on Evolutionary Computing. Workshop in Leeds, UK, April 11-13, 1994 stack machine GP tried it out the results seemed intresting but harder to intepret discusses some work we did using stack and conventional gp.
Uses stack based and conventional GP to fit formulae to describe geographic distance-to-work travel data. cf earlier attempts to do the same using bit string GA. Both GP approaches more successful than simple theoretical model and ``compare well'' with more complex models. Constants agumented by parameters which are optimised outside the GP by nonlinear parameter estimation, high CPU cost but easier for GP. Suggests work needed on good ways to validate GP software.
vector (parallel) processor
",
