The purpose of this work is to learn how to build a neural network, how to set parameters, and to see how the network works with certain settings, and in the end achieve a higher rate od accuracy. Theory begins with the explanation od base terms and basic parts of the network: neurons, the link between neurons, type of network, learning method. All those terms have specific parameters which are chosen in a computer program and are systematically and methodically changing during research implementation. Attained results are satisfying in comparison to previous research, which raises the question, where to use in praxis? To some business subject a high rate of accuracy does not need to be important for business, because, realization of a model of neural network does not necessarily need to be cheap, practical or desirable. The advantages of artificial neural network that we attained is that it was built in a very short period of time and gave a high rate of accuracy. One disadvantage is when it is used for solving problem of classification, but some problems can't be solved. When using these methods it is important that the data being used, be processed, transformed, compatible, consistent etc. which would give the algorithm a much better result. The work elaborates definitely on a wide spectre that enters into many branches of science which can be easily implemented, which encourages a variety of scientist's to contribute and to give their thoughts on the subject matter, and in the end to make a „step forward“.