Hierarchical pattern classifier reduces number of comparisons between input and memory vectors without reducing detail of final classification by dividing classification process into coarse-to-fine hierarchy that comprises first "grouping" step and second classification step. Three-layer neural network reduces computation further by reducing number of vector dimensions in processing. Concept applicable to pattern-classification problems with need to reduce amount of computation necessary to classify, identify, or match patterns to desired degree of resolution.
Hierarchical Pattern Classifier
NASA Tech Briefs ; 16 , 6
1992-06-01
Miscellaneous
No indication
English
Road intersection detection and classification using hierarchical SVM classifier
Tema Archive | 2014
|A Computational Model of a Gazing Point Controlling Process in a Hierarchical Classifier System
British Library Online Contents | 2000
|A Trainable N-tuple Pattern classifier and its Application for Monitoring Fish Underwater
British Library Conference Proceedings | 1999
|