Title: A data-driven prediction of life expectancy trends for 266 countries using time series and polynomial regression models

Authors: G. Paavai Anand; K. Aravind Ram; S. Satyavolu Surya Vamsi Krishna

Addresses: Department of Computer Science and Engineering, SRM Institute of Science and Technology, Vadapalani, Tamil Nadu, India ' Department of Computer Science and Engineering, SRM Institute of Science and Technology, Vadapalani, Tamil Nadu, India ' Department of Computer Science and Engineering, SRM Institute of Science and Technology, Vadapalani, Tamil Nadu, India

Abstract: Life expectancy prediction is vital for policy-making and public health planning. This study applies artificial intelligence to model life expectancy trends across 266 countries using World Bank census data. A hybrid approach combines the auto-regressive integrated moving average (ARIMA) model for capturing linear, time-based trends with polynomial regression to model non-linear residual patterns. The polynomial degree is optimised (1 to 5) to minimise error, enhancing accuracy per country. This dual-model system improves robustness and adaptability to various datasets of the countries, enabling reliable, country-specific predictions. This approach supports informed decision-making for governments and public health bodies worldwide.

Keywords: ARIMA; mean absolute error; MAE; mean squared error; MSE; polynomial regression.

DOI: 10.1504/IJAIH.2025.149247

International Journal of Artificial Intelligence in Healthcare, 2025 Vol.1 No.1, pp.60 - 74

Received: 03 Dec 2024
Accepted: 15 Apr 2025
Published online: 20 Oct 2025 *