This study investigates the relationship between smoking prevalence and life expectancy in Indonesian provinces using data mining techniques, specifically focusing on the application of random forests. The primary objective is to quantify the potential impact of reducing smoking prevalence on population health outcomes. Data were sourced from the Indonesian Central Bureau of Statistics, which included life expectancy and smoking prevalence data from 2021 to 2023. The methodology involved aggregating life expectancy data from the district to the province level, followed by the application of a random forest model to predict life expectancy based on smoking prevalence and other socioeconomic indicators. Key findings indicate a weak to moderate negative correlation between smoking prevalence and life expectancy, with higher smoking rates associated with lower life expectancies. Predictive modeling suggests that a reduction in smoking prevalence could lead to significant improvements in life expectancy. For example, a 5% reduction in smoking rates could increase the average life expectancy by approximately 0.3 years, while a 15% reduction could result in an increase of about 0.9 years by 2025. These results underscore the detrimental impact of smoking on population health and highlight the importance of effective tobacco control measures. The predictive models developed in this study provide valuable information for policymakers, enabling targeted public health strategies and resource allocation. This research contributes to the field by demonstrating the utility of data mining techniques in public health and offering a comprehensive analysis of the relationship between smoking and life expectancy in Indonesia. The findings advocate for the urgent implementation of smoking cessation programs to enhance life expectancy and improve public health outcomes
Copyrights © 2024