Penerapan Metode Interpolasi Polinomial Newton dengan Selisih Terbagi untuk Estimasi Pertumbuhan Jumlah Penduduk di Kota Surakarta

Deshinta Yura Saraswati (Tadris Matematika, Fakultas Ilmu Tarbiyah, Universitas Islam Raden Mas Said Surakarta)
Astutik Apriliana (Universitas Islam Raden Mas Said Surakarta)
Dhina Laras Ati (Universitas Islam Raden Mas Said Surakarta)
Ari Wibowo (Universitas Islam Raden Mas Said Surakarta)

JASIEK (Jurnal Aplikasi Sains, Informasi, Elektronika dan Komputer)
Vol. 8 No. 1 (2026): Juni 2026

Population growth is a key indicator in regional development planning because it affects economic, social, and infrastructure aspects. The city of Surakarta has experienced steady population growth, necessitating the use of accurate estimation methods. This study aims to estimate population growth using the Newton polynomial interpolation method with divided differences and to compare it with linear regression. This study employs a quantitative approach using secondary data from 2021 to 2025 obtained from the Central Statistics Agency of Surakarta City. The analysis was conducted by compiling a split-difference table, forming a Newton polynomial, and creating a linear regression model using Python. The results show that the Newton interpolation method produced population estimates of 540,943 people for 2026 and 583,103 people for 2027, with a MAPE of 0% and an RMSE of 0. Meanwhile, linear regression produced population estimates of 531,259 for 2026 and 533,025 for 2027, with a MAPE of 0.116% and an RMSE of 694.5. Thus, the Newton interpolation method is more appropriate for representing historical data, while linear regression is more appropriate for long-term predictions.