MODELING REGRESSION ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM (RANFIS) FOR PANEL DATA

Panel data combine cross-sectional and time-series data. Data on economic, business, social, development issues are often presented in panel In constructing the regression model, it is necessary to take various steps for testing model specifications, including Chow test Hausman test. This study constructed a classical adaptive neuro-fuzzy inference system (RANFIS). The RANFIS by applying fuzzy neural network (NN) techniques expected overcome problem of uncertainty. One main problems an optimal selecting input variables. variables selected basis best regression. These inputs classified into clusters, which depend degree membership functions. rule bases determined its clusters. empirical this research construct Human Development Index (HDI) Central Java 2017-2019. HDI depends several such as school participation rate, number health workers, public complaints, population growth poverty severity index predictor Based regression, three were used modeling. Evaluation performance was measured based RMSE MAPE values. RANFIS, values 3.227 3.299, respectively. Keywords: Regression, Index, Regression Adaptive Neuro-Fuzzy Inference System DOI： https://doi.org/10.35741/issn.0258-2724.58.3.57