Prediction of Indonesia School Enrollment Rate by Using Adaptive Neuro Fuzzy Inference System

Abstract

The study aimed to predict the school enrollment rate in Indonesia using the Adaptive Neuro Fuzzy Inference System (ANFIS). ANFIS is a combination of fuzzy inference system and artificial neural networks. The study used the Gaussian and Gbell membership functions to make the predictions. The results were evaluated using the R square score (coefficient of determination) and Mean Square Error methods. The results showed that the model performed well in predicting the school enrollment rate, particularly in the age categories of 7-12 years and 13-15 years. The R square score for these categories was 0.981551771 and 0.989081085, respectively, while the Mean Square Error was 0.023947290 and 0.3675162695238, respectively. The performance of the model in the age categories of 16-18 years and 19-24 years was also good, but with a slightly lower R square score and Mean Square Error compared to the younger age categories. When using the Gaussian membership function, the model performed even better, particularly in the age categories of 13-15 years and 19-24 years. The R square score for these categories was 0.99020792 and 0.9883091, respectively, while the Mean Square Error was 0.32958834 and 0.31523466571, respectively. Overall, the study demonstrated that ANFIS is a suitable method for predicting school enrollment rate in Indonesia. The results from this study can provide useful information for decision makers in the education sector, who can use the model to make informed decisions about future educational policies and programs.