Time Series Analysis of Vehicle Accidents in Umuahia (2001 — 2005), Abia State:- Emeruwa Confidence ,N

Abstract

This Project was aimed at carrying out a time series analysis of vehicle accidents in Abia State, a case study of Umuahia from 2001 to 2005. In this research work, I identified the model and fitted the model. On the analysis, it was observed that ACF (autocorrelation function) was differences once to obtain stationarity and it cut-offs at first lag, that is MA(1) (moving average model of order 1). And PAC.f (partial autocorrelation function) were cut-offs at autocorrelation function) were cut-offs at lag 1 and 2, it suggested in autoregressive models of order two, that is AR(2), but fewer lag cuts-ff at lower lags, so 1 chose AR(0) or AR(1). By using Normality probability plots of the residuals of the model, I observed that the ARIMA (0, 1, 1) gave a normal curve implying that the model is fit for the data. The estimates ofthe model parameters given by the MINITAB softwares are 0O = -0.0842 ft = 16.406 To test the forecasting power ofthe fitted model, Error of forecast (f) were calculated for the model, since it has a relatively minimum MSE, we may infer that an unbiased forecast was provided by the model.