Proximal manifold learning via descriptive neighbourhood selection
James Francis Peters, Randima Hettiarachchi · Applied Mathematical Sciences · 2014
This article introduces proximal manifold learning via descriptive neighbourhood selection, which imposes a descriptive form of isometric mapping on high dimensional proximal data spaces into lower dimensional proximal manifolds in descriptive Efremoviyc(EF) proximity spaces. Proximal data spaces are defined by n-dimensional feature vectors in the Euclidean space R n . This leads to results concerning descriptively near manifolds in pictures (digital images), which are useful in discerning picture patterns.