Multivariance of the triangles description in their hierarchical classification and assessing the probability of hitting a target silhouette
Year: 2025
Type: article
Abstract: : This paper addresses the problem of multivariance in the hierarchical classification of triangles, which arises when different traversal rules—differing in the choice of the starting vertex, the resolution of coordinate ambiguities, and the direction of traversal—produce different tree-graph structures and different class assignments for the same set of twenty-four triangle classes. Eight potential traversal variants are identified, of which four are implemented as tree graphs. The study introduces the diagram of permitted directions as a tool for representing allowable movements along triangle sides at each hierarchical level. A forest of tree graphs is proposed as a means to resolve multivariance, where each tree corresponds to a specific traversal rule. A logical verification block is incorporated into the classification procedure to ensure that the correct triangle class is unambiguously determined before the probability estimation step. The probability of hitting a target silhouette is computed by summing the contributions of individual triangles in a Delaunay triangulation network, with each triangle assigned its class according to the selected traversal rule. Keywords: hierarchical classification of triangles, multivariance, taxon, forest of trees, diagram of permitted directions, hitting probability. Contents:IntroductionHierarchical classification of trianglesDiagrams of permitted directionsMultivariance of hierarchical classification of triangles using the feature of traversal directionAlgorithm for estimating hit probability accounting for multivariance in triangle descriptionConclusionReferences Suggested citation: Khaikov, V.L. Multivariance of the triangles description in their hierarchical classification and assessing the probability of hitting a target silhouette. // Electronic Information Systems, No. 2 (45), 2025, pp. 82–97. EDN: GYSROD. (In Russian)
Author: Khaikov, Vadim
Language: Russian
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7208831461