Time series analysis of pneumonia case in Uganda using an Arima mode from 2014 to 2021
MetadataShow full item record
A time series is a sequence of data points, typically measured at uniform time intervals. Examples occur in a variety of fields ranging from economics to engineering, and methods of analyzing time series constitute an important part of Statistics. Time series analysis comprises methods for analyzing time series data in order to extract meaningful characteristics of the data and forecast future values. The Autoregressive Integrated Moving Average (ARIMA) models, or Box-Jenkins methodology, are a class of linear models that are capable of representing stationary as well as nonstationary time series. Pneumonia is one of the deadly neglected disease in Uganda This study was to investigate the times series cases of Pneumonia in Uganda between 2014 to 2021 and the frequency of cases was in months with the objective to see the trend and forecast of cases. The study used the ARIMA model to achieve its Objectives. The analysis used approach of Box–Jenkins’s methodology, where modelling including both finite and infinite lag models to forecast Pneumonia incident cases in the Country. A time series plots of the Pneumonia cases depicted that the series fluctuates with time in an increasing and decreasing. The Autocorrelation and Partial Autocorrelation plots, the proposed candidate models that were tested to be fitted due to TB include ARIMA (2,1,0). The models developed for predicting the monthly Pneumonia cases was adequate for representing the series as evident from all the diagnostics and model comparison techniques employed in the study. However, the forecast, was based on assessment from the linear ARIMA model, Predicted Pneumonia cases. It is noticeable that the series exhibits a trend which is up and down in nature meaning that it is non-stationary A unit root test conducted to investigate the stationarity of the pneumonia cases clearly revealed that the data was not stationary. This was affirmed by the time series plot, ACF and PACF plots of pneumonia cases. The series was then transformed logarithmically and first differenced. From the results, the ADF test revealed that the transformed first differenced series was stationary. The model is a good fit since Prob > chi2 = 0.0000< 0.05 some of the coefficients of the model (AR (2), MA (1)) are significant since their P- values are less than 0.05. It is recommended that the MoH should collaborate with health personnel to provide intensive education on some of the dangers of the disease and the need to seek early treatment in any nearby health facility and that there should be organized effort to mobilize communities on health benefits and use kitchen with enough smoke escape roots like windows and/or chimneys.