For every new sample :
Initially every sample is a cluster.
Repeat until the desired number of clusters is obtained :
Randomly chose k clusters.
For every sample :
A variant that initializes centers so that there is a guarantee in accuracy, and a faster convergence :
A way to chose the optimal number of clusters k.
Compute for different number of clusters the ratio of the intra-clusters variance to the total variance. The optimal number of clusters is when adding clusters do not bring significant decrease of the ratio.
Cartes auto-organisatrices de Kohonen