Three extensions of the traditional learning rule for Self-Organizing Maps are presented. They are based on geometrical considerations and explore various possibilities regarding the norm and the direction of the adaptation vectors. The performance and convergence of each rule is evaluated by two criteria: topology preservation and quantization error.
Lee, J., Donckers, N., & Verleysen, M. (2001). Recursive learning rules for SOMs. In N. Allinson, H. Yin, L. Allinson, J. Slack (ed.), Advances in Self-Organizing Maps (p. p. 67-72). Springer Verlag. https://hdl.handle.net/2078.5/253811