LitMy.ru - литература в один клик

Intelligent Multimedia Data Analysis

  • Добавил: literator
  • Дата: 24-02-2020, 06:56
  • Комментариев: 0
Название: Intelligent Multimedia Data Analysis
Автор: Siddhartha Bhattacharyya, Indrajit Pan
Издательство: De Gruyter
Год: 2019
Страниц: 197
Язык: английский
Формат: pdf (true)
Размер: 10.1 MB

Multimedia encompasses varied forms of media information which can be classified into discrete media (textual and image) and continuous media (animation, audio and video). Multimedia data often initiate from human communication, thus contain uncertainties and ambiguities. The book describes intelligent tools and techniques to handle such uncertainties and ambiguities. Examples include neural networking, fuzzy logic and the evolutionary computing.

Soft computing technologies are quite efficient at handling the imprecision and uncertainty of multimedia data and they are flexible enough to process real-world information. In spite of the several advantages of soft computing techniques, many times soft computing tools and techniques are unable to provide a solution to problems. As a result, different types of hybridization of soft computing techniques such as neuro-fuzzy, neuro-genetic, fuzzy-genetic, neuro-fuzzy-genetic, and rough-fuzzy, etc. are proposed to provide more robust and failsafe solutions. Hybrid soft computing stems from the synergistic integration of the different soft computing tools and tech-
niques. The fusion of these techniques to enhance performance and produce more robust solutions can be achieved through proper hybridization. An intelligent machine inherits the boon of intelligence by virtue of the various methodologies offered by the soft computing paradigm encompassing fuzzy and rough set theory, artificial neurocomputing, evolutionary computing, as well as approximate reasoning.

Скачать Intelligent Multimedia Data Analysis












[related-news] [/related-news]
Внимание
Уважаемый посетитель, Вы зашли на сайт как незарегистрированный пользователь.
Мы рекомендуем Вам зарегистрироваться либо войти на сайт под своим именем.