This paper presents an unsupervised segmentation method for feature sequences based on competitivelearning hidden Markov models. Models associated with the nodes of the Self-Organ...
The task of finding the optimum of some function f(x) is commonly accomplished by generating and testing sample solutions iteratively, choosing each new sample x heuristically on t...
One of the most challenging issues in managing the large and diverse data available on the World Wide Web is the design of interactive systems to organize and represent information...
Designing and interoperability testing of distributed, application-level network protocols is complex. Windows, for example, supports currently more than 200 protocols, ranging fro...
Recently we have proposed an algorithm of constructing hierarchical neural network classifiers (HNNC), that is based on a modification of error back-propagation. This algorithm co...
S. A. Dolenko, Yu. V. Orlov, I. G. Persiantsev, Ju...