A list of interesting concepts, ideas that I stumbled upon at some point in time

**convolution vs correlation**: they differ only in the ‘flipping’ of the filter, see e.g. Convolution. That means that for symmetrical filter functions there is actually no difference. On the other hand, this answer points out that the interesting point in convolutions is that they are associative and therefore are interesting when multiple filters should be composed (*2018-02-11*).**Relative neighborhood graph**: inspired to show the structure that a human eye perceives in a set of points. See Triangulation for a playground (*2018-01-27*)**Abilene paradox**: when group decisions go wrong despite best intentions (*2018-01-27*)**quadtree / spatial hashing**: two methods/data structures used to store objects while keeping their spatial relationship to some extent, allowing more effective calculations, e.g. for collision detection - both are described in this comparison (*2017-12-30*)**Delaunay triangulation**as dual to Voronoi tessellation: The Wikipedia article explains connection between them. See Triangulation for a playground (*2017-12-30*)**auxetic materials**are technically defined through a negative Poisson’s ratio - meaning for example that they get thicker while stretched!- Generative Adversarial Networks (GAN)
- autoencoders
- context-free grammars

2018-02-23