| Shanice Douglas |
Lung and colon cancers remain among the top causes of cancer-related mortality globally, with significant variations
across health systems. This study applies multivariate time series modeling, specifically the Vector Autoregressive
Moving Average (VARMA) approach, to analyze lung and colon cancer mortality trends in Jamaica and the United States
from 1960 to 2021. It explores interdependencies between the two cancer types within each country and provides 12-year
ahead forecasts. Model performance is evaluated using AIC, RMSE, and MAPE, with Monte Carlo simulations enhancing
forecast reliability. A comparative assessment highlights cross-country disparities, shared patterns, and potential
contributing factors. Findings aim to inform predictive analytics and guide public health policy for better cancer
control.
Keywords: Cancer Mortality, Multivariate Forecasting, VARMA, Time Series, Monte Carlo Simulation |