doktorska disertacija
Iztok Fister (Author), Janez Brest (Mentor)

Abstract

Algoritmi računske inteligence so metode, ki delujejo po vzorih iz narave in poskušajo reševati težke probleme s posnemanjem principov naravnih sistemov. Med te metode v grobem štejemo: nevronske mreže, evolucijske algoritme, algoritme inteligence rojev, umetne imunske sisteme, sisteme na osnovi mehke logike in verjetnostne metode. Skozi zgodovino so se ti algoritmi uspešno uporabljali za reševanje problemov na skoraj vseh področjih človekovega udejstvovanja, vendar do nedavnega njihove prisotnosti ni bilo zaznati na področju športa. Doktorska disertacija tako odpira novo raziskovalno področje, kjer algoritme računske inteligence uporabimo za razvoj umetnega športnega trenerja. Umetni športni trener je sistem, ki omogoča vključevanje algoritmov računske inteligence za podporo različnih faz šport\-nega treninga. V prvem delu doktorske disertacije naredimo pregled obstoječih algoritmov računske inteligence, se dotaknemo osnov športnega treninga in orišemo koncept umetnega športnega trenerja. V eksperimentalnem delu doktorske disertacije predstavljamo praktična primera uporabe umetnega šport\-nega trenerja. Prvi primer prikazuje načrtovanje športnih treningov za različne časovne cikle, medtem ko drugi vključuje uporabo algoritma rojne inteligence za odkrivanje navad športnikov. Pridobljeni rezultati dokazujejo učinkovitost umetnega trenerja ter vzpodbujajo njegov nadaljnji razvoj.

Keywords

algoritmi računalniške inteligence;inteligenca rojev;načrtovanje športnih treningov;podatkovno rudarjenje;umetni športni trener;doktorske disertacije;

Data

Language: Slovenian
Year of publishing:
Typology: 2.08 - Doctoral Dissertation
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: [I. Fister ml.]
UDC: 004.421.2:004.8(043.3)
COBISS: 20795670 Link will open in a new window
Views: 1868
Downloads: 395
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Computational intelligence algorithms for the development of artificial sport trainer
Secondary abstract: Computational intelligence algorithms are methods inspired by nature and, thus, they try to solve hard problems by mimicking the behavior of natural systems. Under a common umbrella, this algorithm family associates the following aforementioned members: artificial neural networks, evolutionary algorithms, swarm intelligence algorithms, artificial immune systems, fuzzy systems and probability methods. These algorithms have been applied successfully to almost all areas of human engagement through time. However, their presence within the sport domain has not been perceived yet. Therefore, this dissertation opens a new research area, where computational intelligence algorithms are used for the development of an Artificial Sport Trainer. Roughly speaking, an Artificial Sport Trainer is a system that involves computational intelligence algorithms in order to support different stages of sport training. In the first part of the existent dissertation, we outline an overview of existing computational intelligence algorithms, touch the fundamentals of sport training, and present the concept of an Artificial Sport Trainer. In the experimental part of the dissertation, we present two practical applications of an Artificial Sport Trainer. The former presents planning the sport sessions for various sport training cycles, while the latter deals with the discovering of an athlete's habits during sports training sessions. The obtained results prove the efficiency of the Artificial Sport Trainer and encourage us for further development.
Secondary keywords: artificial sport trainer;computational intelligence algorithms;data mining;sport session planning;swarm intelligence;
URN: URN:SI:UM:
Type (COBISS): Doctoral dissertation
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko
Pages: XVI, 155 str.
ID: 10858945