论文信息 - Analysing a Contingency Table with Kohonen Maps: A Factorial Correspondence Analysis

Analysing a Contingency Table with Kohonen Maps: A Factorial Correspondence Analysis

The Kohonen self-organizing algorithm is a powerful tool to achieve a categorization of vectorial stochastic data into classes. Many researchers use it to get a preliminary reduction of the data complexity in numerous application fields. They address some problems which are usually solved by means of statistical methods like Classification, or Principal Component Analysis. In this paper, we propose to extend this approach to another data analysis method: the simultaneous analysis of two qualitative variables which are crossed in a contingency table.

[1] Teuvo Kohonen,et al. The self-organizing map , 1990 .

[2] R. Palmer,et al. Introduction to the theory of neural computation , 1994, The advanced book program.

[3] Pierre Demartines,et al. Data Analysis: How to Compare Kohonen Neural Networks to Other Techniques? , 1991, IWANN.

[4] Helge J. Ritter,et al. Neural computation and self-organizing maps - an introduction , 1992, Computation and neural systems series.

[5] G. Pagès,et al. Self-organization and a.s. convergence of the one-dimensional Kohonen algorithm with non-uniformly distributed stimuli , 1993 .

[6] Teuvo Kohonen,et al. Self-Organization and Associative Memory , 1988 .

[7] P. Duncombe,et al. Multivariate Descriptive Statistical Analysis: Correspondence Analysis and Related Techniques for Large Matrices , 1985 .

[8] Gilbert Saporta,et al. L'analyse des données , 1981 .