The Assessment and Comparison of Performance Analysis Using Exponentially Weighted Moving Average and Cumulative Sum Scheme.

Abstract

This research work is designed for monitoring some prominent road traffic offences in Kwara state of Nigeria .A historical value of the most prominent road traffic offences were used to construct the prediction interval using the non-parametric and bootstrap methods at two different significant levels. The prediction intervals obtained can be used as tool to identify future values that suggest changes in the frequencies the road traffic offences considered in the study. The interval can be used to predict changes in the future values. The two approaches are compared and random sample of sizes 35 and 25 from the studied data were used to evaluate the robustness in variation of the two approaches. The results indicate that the two approaches exhibit similar patterns. However, the bootstrap approach prediction interval is more stringent than the non-parametric approach.