Κατασκευή τεχνητών νευρωνικών δικτύων με τη χρήση του Wolfram Mathematica
Construction of artificial neural networks using Wolfram Mathematica
View/
Keywords
Τεχνητά νευρωνικά δίκτυα ; Wolfram Mathematica ; Ταξινόμηση εικόνων ; Χρωματισμός εικόνων ; CNN ; ANN ; Image classification ; Image colorizationAbstract
This master thesis explores the construction and application of artificial neural networks using the Wolfram Mathematica software. Initially, it provides an introduction to neural networks, presenting their structure and function, both biologically and artificially. This
section analyzes various activation functions and describes training methods of artificial neural networks, their architectures, and the different models used. The thesis then focuses on creating and improving neural networks using Wolfram Mathematica. It presents the fundamental building blocks of the networks, such as the various layers, and analyzes the commands used for constructing these networks. A significant part of the thesis is dedicated to applications of neural networks in Wolfram Mathematica. Specific application examples of these networks are presented, such as image classification and colorization of grayscale images. Specifically, the CIFAR-10 dataset is used for image classification, while the Image Colorization Dataset is used for colorization of grayscale images.The thesis provides a comprehensive view of the use of Wolfram Mathematica in the
construction and training of artificial neural networks, highlighting the capabilities and challenges of this approach.
Postgraduate Studies Programme
Εφαρμοσμένη ΣτατιστικήDepartment
Σχολή Χρηματοοικονομικής και Στατιστικής. Τμήμα Στατιστικής και Ασφαλιστικής ΕπιστήμηςNumber of pages
114Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα