Dissimilarity-Based Correlation of Movements and Events on Circular Scales of Space and Time

Abstract

© 2020, Springer Nature Switzerland AG. Circular scales appear in many applications related to a comparative analysis of the timing of events, wind directions, animals and vehicle movement directions, etc. The paper introduces a new non-statistical correlation function on circular scales based on a recently proposed approach to constructing correlation functions (association measures) using (dis)similarity measures on the set with an involutive operation. An involutive negation and a dissimilarity function satisfying required properties on a circular set are introduced and used for constructing the new correlation function. This correlation function can measure correlation both between two grades of circular scale and between sets of measurements of two circular variables.