Forecasting the Real GDP in Nigeria Using Possibilistic Linear Regression Model

Abstract

This paper presents application of fuzzy linear regression to forecast the real GDP of Nigeria based on macroeconomic indicators including unemployment rate, inflation rate and FDI. A fuzzy linear regression model capable of predicting the real GDP of Nigeria, assuming that residuals are due to system fuzziness is ideal. Based on the empirical results, two fuzzy regression models with threshold values of 0 and 0.5 that can adequately estimate the real GDP of Nigeria are established. Consequently, considering the criteria of interval of possibility and MAPE, the most suitable model is determined.