Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Forecasting unemployment rate in Greece by the neuro fuzzy system Anfis

Kalpakidou Maria

Full record

---

Summary

The purpose of this study is to present the possibility of predicting unemployment rates in Greece, taking data from 2004-2018, using the adaptive neuro fuzzy system /ANFIS. The Anfis system is been selected after bibliographic research with other methods of forecasting in other research areas. Compared to other methods, it has high success rates and demonstrates its superiority and opportunities for evolution to the adaptive nerve fuzzy networks, which are been tested in the present study at unemployment rates. The phenomenon of unemployment raises intense concerns and different theoretical approaches. Searching for information from the past and processing it with appropriate predictive methods enables us to gain the necessary knowledge for the present and the future. The main objective is to design active policies with medium and long-term benefits to society. In the introduction is been presented a definition of unemployment, the types of measurement, the causes, the consequences, and the ways of dealing with it. After, there is a reference to the prediction methods, presentation in summary, the concept of fuzzy logic, function of participation, properties of fuzzy sets, logical actions in fuzzy sets, Below, we mention the neural networks definition, historical data, artificial neural networks, neural network use and architectural neural network applications, types of architectural structures. Next, there is an introduction to the theory of adaptive neurodegenerative systems with extensive reference to the Anfis model and its architecture. Afterwards, the case study of unemployment rates been analyzed, programming in Matlab environment and the results of the model. Finally, we come to conclusions about the forecasting abilities of Anfis system on unemployment rates.

Available Files

Services