1Hamidu Aliyu Chamalwa and 2Faisal Mohammed Kumshe
1Department of Statistics, Faculty of Physical Sciences, University of Maiduguri-Borno State
2Department of Mathematics and Computer Science, Faculty of Science, Borno State University, Maiduguri-Borno State
*Corresponding author’s Email: chamalwa@unimaid.edu.ng, doi.org/10.55639/607.232221
ABSTRACT
The research utilizes the ARIMA model to conduct a comprehensive time series analysis on infant mortality rates in Maiduguri, revealing trends and seasonality. For both genders, there’s a recent fluctuation in trends. Autocorrelation and partial autocorrelation functions are employed to determine the model’s order. The data were transformed into stationary by differencing once after which ADF and KPSS were used to confirm the stationary status; several tentative models were fitted and assessed for best fit; using the AIC, BIC and HQIC criteria. The chosen models are ARIMA (0,1,2) for male and ARIMA (0,1,1) for female was validated through diagnostics and employed for forecasting. Forecasts for 2024 indicate an unstable trend for males and stability for females. Recommendations include: enhancing delivery channels and community-based services, particularly in high-mortality areas, emphasizing girl-child education through religious and traditional leader advocacy campaigns.
Keywords:
Assess,
Child,
Death,
Fitted,
Order