Fourier series nonparametric regression estimator for modeling status of unmet need in East Java Province in 2023

Abstract

In recent years, the Estimator of Fourier Series Nonparametric Regression (FSNR) for quantitative data has generated a lot of attention. In practice, though, there is frequently a correlation between predictor and response, with categorical data serving as the response. Only certain techniques are used in some of the methodologies created today to address the health case of qualitative response data. No previous study can handle health data using FSNR estimator. This study presents a novel approach FSNR estimator with response variables in the form of categorical data specifically within the context of public health research. The research methods used are theoretical and application studies. The FSNR estimators method for categorical data assumes a relationship between the logit function and predictor variables that has a repeating pattern. The Newton-Raphson technique and MLE were used to obtain the FSNR estimators. To apply this method, we used application data status of unmet need in East Java Province in 2023. The unmet need rate in East Java was quite high, reaching 12.97 percent, while the target is 11.74 percent. Because of the lower deviance value, greater AUC, and Press's Q values, the results show that the FSNR offers much superior estimation results and accuracy for data applications.