Can Neural Networks Be Used to Define the Rules of Cardiovascular Disease Prevention in the Nutrition Domain?

  • Agnieszka Borowiec
  • Krzysztof Fronczyk
  • Bohdan Macukow
  • Maciej Grzenda
Conference paper
Part of the Advances in Soft Computing book series (AINSC, volume 19)


The primary objective of the work was to check the relation between household socio-economic characteristics and the corresponding food purchasing capabilities. The analysis has been based on the data collected in the national survey “Social Diagnosis 2000”. In order to search for possible dependencies between variables gathered in the survey, different classification methods have been applied.

Statistic methods of logistic regression analysis and discriminant analysis have been applied to model the discussed relation. However, only limited prediction efficiency has been observed. Therefore, neural networks-based methods have been applied. Evolutionary construction of multilayer perceptrons has been used to select both network architectures and weights. The method developed and tested previously on numerous prediction and classification problems has been used to provide classification models for the data discussed. Multilayer perceptrons have been shown to provide more precise classification models. Results of the network construction are presented together with final discussion.


Multilayer Perceptrons Data Pattern Food Category Load Forecast Social Data 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bäck, T., Schwefel, H.P. (1993) An Overview of Evolutionary Algorithms for Parameter Optimization. Evolutionary Computation. MIT Press,Vol. 1 No. 1, 1–23Google Scholar
  2. 2.
    Bäck, T. (1998) An Overview of Parameter Control Methods by Self-Adaptation in Evolutionary Algorithms. Fundamenta Informaticae. 35, 51–66MATHGoogle Scholar
  3. 3.
    Duch, W., Grabczewski, K. (1999) Searching for Optimal MLP. Proc. of the Fourth Conference Neural Networks and Their Applications. Zakopane, 65–70Google Scholar
  4. 4.
    Fogel, D.B. (1999) An Overview of Evolutionary Programming, The IMA Volumes in Mathematics and its Applications, ed. Williard Miller, 111, 89–109Google Scholar
  5. 5.
    Grzenda, M., Macukow, B. (2001) Evolutionary Model for Short Term Load Forecasting. Proc. of MENDEL 2001 conference. Brno, 119–124Google Scholar
  6. 6.
    Haykin, S. (1999) Neural Networks: a Comprehensive Foundation, Prentice-Hall Inc., USAGoogle Scholar
  7. 7.
    Hosmer, D.W., Lemeshow, S. (1989) Applied Logistic Regression, John Wiley Sc Sons, USAGoogle Scholar
  8. 8.
    Klecka, W.R. (1980) Discriminant Analysis, Sage, LondonGoogle Scholar
  9. 9.
    Macukow, B., Grzenda, M. (2001) Towards the Evolution of Neural Networks. Opto-Electronics Review, 9 (3), 316–319Google Scholar
  10. 10.
    Siwek, K. (2000) Load Forecasting Using Artificial Neural Networks, Phd thesis, Warsaw University of Technology (in Polish)Google Scholar
  11. 11.
    Stamler, J. (1999) Nutrition and Longevity: Multifaceted Improvements in Population Dietary Patterns to End the Cardiovascular Disease Epidemic. Nutrition, metabolism and cardiovascular diseases. 9(4) Suppl, 37–51Google Scholar
  12. 12.
    Ross, E. et al. (1999) Gender, Socioeconomic Status and Family Status as Determinants of Food Behaviour. Social Science and Medicine. Vol. 46 No. 12, 1519–1529Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Agnieszka Borowiec
    • 1
  • Krzysztof Fronczyk
    • 1
  • Bohdan Macukow
    • 2
  • Maciej Grzenda
    • 2
  1. 1.Department of Health PromotionNational Institute of CardiologyWarszawaPoland
  2. 2.Faculty of Mathematics and Information ScienceWarsaw University of TechnologyWarszawaPoland

Personalised recommendations