A pattern classification scheme in which the classifier is able to grow and evolve during the operation process is presented. The evolutionary property of the classifier is made possible by modeling the pattern vectors in multiple hyper-ellipsoidal subclass distributions. Learning of the classifier takes place at the subclass levels only. This property allows the classifier to retain its previously learned patterns while accepting and learning new pattern classes. The classifier is suitable to operate in dynamical environments where continuous updating of the pattern class distributions is needed.
ASJC Scopus subject areas
- Signal Processing
- Computer Vision and Pattern Recognition
- Artificial Intelligence