Dealing efficiently with huge amounts of high-dimensional data requires a quick and reliable, but at the same time also robust information storage and retrieval. From this perspective, we study various models of self-organizing feature maps.