Journal of Engineering and Engineering Technology - JEET |ASSESSING THE HYBRID OF ARIMAX AND ANN MODELS FOR PREDICTING LIFE EXPECTANCY IN NIGERIA

Paper Details

Title :ASSESSING THE HYBRID OF ARIMAX AND ANN MODELS FOR PREDICTING LIFE EXPECTANCY IN NIGERIA
Author(s) : Christopher Awariefe (1) and Godwin Ekruyota (2)
DOI:
Abstract:

This research investigates the relationship between life expectancy and key variables in Nigeria, employing advanced forecasting models to address a research gap in understanding the collective influence of economic and demographic factors on life expectancy. The study focuses on the Hybrid ARIMAX-ANN method and compares its performance to ARIMA, ARIMAX, and ANN models. Data from 1960 to 2022 are utilized, with auto.arima() for ARIMA model selection, ARIMAX incorporating exogenous variables, and ANN employing a feed-forward neural network. The research emphasizes accurate model selection for reliable forecasting. Results show that the Hybrid ARIMAX-ANN model outperforms other models, demonstrating superior accuracy in predicting Nigerian life expectancy. Performance metrics such as MAPE and RMSE confirm the effectiveness of the hybrid model. Visual comparisons and forecast accuracy evaluations of the train data further support these findings. The study contributes to the predictive modelling discourse for life expectancy in developing nations, providing insights into the interplay of economic and demographic factors in Nigeria. These insights are expected to guide evidence-based policymaking and public health interventions, leading to improvements in population health outcomes in Nigeria and potentially offering strategies for similar contexts globally.

Pages

Contact us

Email us

Call Us