Melnikov V. Parallel-hierarchical network learning and self-learning methods and tools

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0412U006883

Thesis Registration Form

0412U006883.pdf

Applicant for

Melnikov Victor Volodymyrovych

Specialization

05.13.05 - Комп'ютерні системи та компоненти

Date of defense

18-12-2012

Specialized Academic Board

К 26.820.04

Essay

The thesis is devoted to actual scientific problems of developing new methods and computer tools for image recognition and parameters prediction to improve images handling and identification. The PH-based network population images coding method. On the basis of the improved method pretreatments images and developed learning methods, PH-network and forecasting method using PH-networks developed a lot of computational structures and algorithms to improve the efficiency of further processing, Image identification and parameters prediction. Designed software and hardware platform used GPU.

Thesis supervisor

Timchenko Leonid Ivanovych

Official opponents

Цмоць Іван Григорович
Мартинюк Тетяна Борисівна

Files

aref.doc

dis.doc

Similar theses

0524U000048

Anna V. Khakhanova

Federated computing of vector-matrix transactions in cyber-social systems

0424U000015

Oleksandr Melnyk

Methods and means of forming graphic primitives on a hexagonal grid

0422U100014

Saprykin Oleksandr Sergiyovych

Models for automated analysis and diagnosis of polymorphic viruses in computer systems and networks

0421U103964

Uzdenov Taras Amurovych

Task scheduling methods for GRID-systems with inalienable resources

0521U101938

Mozhaiev Mykhailo O.

Models and methods of synthesis of a specialized computer system in providing forensic science services for justice