Algoritmy automatických obchodních systémů

Abstract

The goal of this thesis is the implementation of automated trading system, which uses trading strategies based on predicting the future development of stock prices by neural networks. In its theoretical part, this thesis compares the existing algorithmical trading systems and explains different approaches to analysing financial data. A part of this thesis is dedicated to explaining the basic theory behind neural networks and evolutionary algorithms. The practical portion of the thesis describes, documents and tests the created trading algorithms.