We propose in this paper a new learning algorithm which uses a probabilistic formalism of topological maps. This algorithm approximates the density distribution of the input set with a mixture of normal distributions. We show that, under certain conditions, the classical Kohonen algorithm is a specific case of this algorithm, therefore allowing a probabilistic interpretation of Kohonen maps.
CITATION STYLE
Anouar, F., Badran, F., & Thiria, S. (1996). Topological maps for mixture densities. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1112 LNCS, pp. 833–838). Springer Verlag. https://doi.org/10.1007/3-540-61510-5_140
Mendeley helps you to discover research relevant for your work.