Using Artificial Neural Networks to Simulate the Financial Stability of Enterprise

Abstract

The financial sustainability of enterprise is an extremely important concept in a market environment. In view of the growing need of society for machine-based data processing to support decision-making, there is a need to find and develop mathematical methods that can meet the task of classifying enterprises into financially stable and financially unstable. The article substantiates the feasibility of applying financial stability simulations using artificial neural networks (ANN), and describes the basic principles for the work of ANN. It also provides an example of building an ANN to accomplish this task, which includes a number of financial indicators, using which makes sense for the simulation of financial stability, the parameters for building the neural network described, choosing the best architecture and method for training the network. The results of the model’s work, which show that artificial neural networks can reasonably accurate classify enterprises by the types of «potential bankrupts» and «stable», have been provided.

Keywords

• financial stability
• artificial neural networks
• perceptron
• bankruptcy
• economic-mathematical modelling
• classification task