Synthesis and Analysis of Classifiers Based on Generalized Model of Identification

Tatur, M. M.; Adzinets, D. N.; Лукашевич, М. М.; Bairak, S.

doi:10.1007/978-3-642-12433-4_62

Cited by 3 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, it is possible to obtain mathematical expressions for the discriminant (sharing) surfaces, optimal from a statistical point of view. " The authors fully share this view, and in their works [5,6], propose to combine the linear discriminant function with elements of fuzzy logic. Thus, an opportunity of recording of a priori data about the task in a generalized model of identification is provided.…”

Section: A Mathematical Model Of Identification (Z-model)mentioning

confidence: 97%

See 1 more Smart Citation

Problem-oriented parallel processors for the solving of classification tasks

Adzinets

Tatur²

2013

The International Conference on Digital Technologies 2013

View full text Add to dashboard Cite

This paper describes in detail, the workings of a prototype version of a parallel processor for high-performance computing of large-scale tasks for identification (classification). The use of such processors will provide the required computing power with a guaranteed quality of decision making. Program settings and architecture cascading, as well as the knowledge of the dependence of performance of hardware resources, will reduce the time frame for the development of intelligent systems for pattern recognition as integrated software and hardware, presented in real-time.

show abstract

Section: A Mathematical Model Of Identification (Z-model)mentioning

confidence: 97%

“…4. [5,6]. In general, the solution refers to the SIMD-architecture (Single Instruction Multiply Data) and is characterized by the following main properties.…”

Section: A Mathematical Model Of Identification (Z-model)mentioning

confidence: 99%