Using holographic neural networks for currency exchange rates prediction

In this paper we describe a series of experiments, whose objective was to evaluate holographic neural networks as a possible tool for currency exchange rates prediction. In our experiments we trained and tested holographic networks on authentic data from "Zagrebacka banka". The results presented in the paper give some indications when neural networks in general are suitable for prediction problems. Also, the results demonstrate certain advantages of holographic vs. traditional neural networks.