The role of Fisher information in primary data space for neighbourhood mapping
Hectór Andrés RUIZ, Ian H. Jarman, José D. Martín‐Guerrero, Paulo Lisböa · 2011
Abstract. Clustering methods and nearest neighbour classifiers typically compute distances between data points as a measure of similarity, with nearby pairs of points considered more like each other than remote pairs. The distance measure of choice is often Euclidean, implicitly treating all directions in space as equally relevant. This paper reviews the application of Fisher information to derive a metric in primary data space. The aim is to provide a natural coordinate space to represent pairwise distances with respect to a probability distribution p(c|x), defined by an external label c, and use it to compute more informative distances. 1