A Time Series Forecasting of the Philippine Unemployment Rate Using Feed-Forward Artificial Neural Network

Unemployment is considered as one of the major sources of social problems and it remains to be a significant challenge to every country. Hence, forecasting the trend of unemployment rate contributes to alleviating a country’s unemployment problem. This study focuses on forecasting the trend of the Philippine unemployment rate using one of the types of architecture of neural network which is the Feed-forward Artificial Neural Network. Neural networks are modern statistical tools. Nowadays these are widely used in different researches because of its ability to process complex and nonlinear data sets. To generate the Philippine unemployment rate forecast, this study used twelve variables namely, Unemployment Rates, Population, Labor Force, Gross Domestic Product (GDP), Gross National Income (GNI), Gross Domestic Investment (GDI), Inflation Rate, Elementary Level Cohort Survival Rate, High School Level Cohort Survival Rate, Higher Education Graduates, Index of value of production of key manufacturing enterprises by Industry and Foreign Trade covering the year 1991-2014 obtained from Philippine Statistics Authority (PSA) Region X. Results show that the model obtained in this study for forecasting the trend of  the unemployment rate in the Philippines is .875 or 87.5% accurate. A mathematical model for forecasting the unemployment rate was also formulated which can be used to generate future estimated values of unemployment rates.

Keywords: Time series forecasting, unemployment rate, feed-forward artificial neural network