Financial Forecasting using Neural Networks

Abstract

Inspired by the human brain structure and functions, artificial neural networks (ANNs) have a large application area in forecasting the economic phenomena. In the actual context, when the economic crises affect the entire business world, it is necessary that financial tools and application software cooperate in order to forecast the market trends with the aim to reduce the insecurity and provide solid growth for the companies in the future.The aim of this paper is the theoretical presentation of a forecasting model based on neural networks and a study case regarding the evolution of the company OMV PETROM S.A., one of the largest companies in Southern and Eastern Europe. For the experiment we used various implementations of the neural network algorithm offered by GMDH Shell software and the financial consolidate statements of the company during 2005-2011, in order to obtain a forecasting of financial indicators for 2012.
Keywords: Corporate management, Financial indicators, Forecasting software, Neural networks.