Fondements biologiques pour le calcul distribué, numérique et adaptatif

6 We present in this paper a variation of the self-organising map algorithm where the original 7 time-dependent (learning rate and neighbourhood) learning function is replaced by a time8 invariant one. This allows for on-line and continuous learning on both static and dynamic 9 data distributions. One of the property of the newly proposed algorithm is that it does 10 not fit the magnification law and the achieved vector density is not directly proportional 11 to the density of the distribution as found in most vector quantisation algorithms. From a 12 biological point of view, this algorithm sheds light on cortical plasticity seen as a dynamic 13 and tight coupling between the environment and the model. 14