Využití neuronových sítí ke kompresi dat z měřicího zařízení

Abstract

The bachelor thesis aims to verify a non - traditional method of data compression, using neural networks. This method offers a different perspective on solving the problem. Thesis itself is divided into two main parts. The theoretical part contains a description of the principles used in neural networks and data compression. The practical part uses the knowledge from the research to design your own network in the Visual Studio Code in Python. The last part of thesis is devoted to experiments that tested the network under various conditions. Accuracy of data reconstruction using RMSE was evaluated. Resulting accuracies were evaluated and compared.