Komprese pomocí neuronové sítě

Abstract

This work deals with data compression using neural networks and compares the advantages of this solution compared to other compression methods. Presents, what the neural networks are and the principles on which they are based. It examines the relation between data compression and neural networks. It is also presenting a basic idea of the other principles of learning neural networks and different compression methods without detailed description. Represents the probability estimate of the bit sequence in reducing the amount of data. Its main objective is to calculate the entropy file size after compression with a focus on text and bitmap formats.