Background Disease forecasting is useful for effective planning of prevention and control program. Time series analysis is one of quantitative methods applied in disease forecasting. Motorcycle injury was a major problem in Thailand. It was reported each year 54000–61000 cases and cases, including 3000–4000 deaths. Most of severe injury case under 15 years old was 5,000–6,000 and 200 died annually. Objective of the study is to compare accuracy between the Exponential Smoothing (ES) and Autoregressive Integrated Moving Average (ARIMA) Models for short-term forecasting of motorcycle injury, in order to estimates of the magnitude of problems, and prepare resources for injury prevention and control.
Methods Monthly data of motorcycle injury numbers from 2006–2014 were collected from Thailand National Injury Surveillance System. Two forecasting methods with the criterion of minimum Mean absolute error (MAE) and Mean Absolute Percent Error (MAPE) based on 2006–2010 data. The selected ES and ARIMA models then were applied for forecasting number of injury in 2011–2014. MAE and MAPE of one-, two-, three-, and four-month forecasting were compared between the two models. P value from paired t-test was calculated for each comparison.
Results The result showed that, in terms of forecasting accuracy, Exponential Smoothing (Simple Seasonal) model performed better than ARIMA (1,0,0) (1,1,1) model. MAPE of the forecasts from ES at one-, two-, three-, and four-month were 2.7%, 6.1%,7.5% and 7.9% respectively, while those from ARIMA methods were 3.0%, 6.4%, 7.8% and 8.3% respectively.
Conclusions We suggested that the Exponential Smoothing (Simple seasonal) should be used as a tool to provide affordable and reliable short-term forecast of motorcycle accident.
- motorcycle injury
- Exponential smoothing
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.