Feedforward Neural Network Model in Unemployment Rates Forecasting in Public Administration
Kouziokas, G. N. (2017). Feedforward Neural Network Model in Unemployment Rates Forecasting in Public Administration. Proceedings of the 6th International Symposium and 28th National Conference on Operational Research, pp. 168-172, June 8-10, 2017, Thessaloniki, Greece
Posted: 31 Jan 2020
Date Written: June 30, 2017
Abstract
The application of Artificial Neural Networks in many scientific fields has been increased the last years with the development of new neural network technologies and techniques In this study, Artificial Neural Networks are applied for building forecasting models in order to predict unemployment. A Feedforward Neural Network structure was used since it is considered as the most suitable in times series predictions. In order to develop the best artificial neural network forecasting model, several network topologies were examined regarding the number of the neurons and also the transfer functions in the hidden layers. Several economic factors were taken into consideration in order to construct the neural network based prediction models. The results have shown a very precise forecasting accuracy regarding the unemployment. The proposed technique can be very helpful in public administration at adopting proactive measures for preventing further increase of unemployment.
Keywords: Artificial Intelligence, Economic Development, Neural Networks, Public Administration, Unemployment
Suggested Citation: Suggested Citation