Comparison of Nadaraya-Watson and Local Polynomial Methods in Analyzing the Relationship Between Consumer Price Index and Inflation in South Kalimantan

Abstrak

This study aims to compare the performance of Nadaraya-Watson and Local Polynomial regression methods in analyzing the relationship between the Consumer Price Index (IHK) and inflation in South Kalimantan Province. Given the potential non-linear nature of this relationship, nonparametric regression approaches were employed as they offer more flexibility compared to traditional parametric models. Data from the Central Statistics Agency (BPS) from January 2022 to December 2024 were used, and missing values in the inflation variable were handled using mean imputation. The optimal bandwidth for both methods was determined using the direct plug-in method (dpill) to ensure consistent comparison. The results show that the Nadaraya-Watson method produces a more fluctuating curve, indicating high sensitivity to local data variations but also a higher risk of capturing noise. In contrast, the Local Polynomial method yields smoother and more stable curves, better capturing the overall trend without being overly affected by local fluctuations. The findings suggest that the choice of method should be based on the analysis objective, whether to emphasize local detail or overall trend.