The benefit of conducting Sliding Window Analysis is that the window position is not fixed and therefore, the accuracy of hotspots identified is maximized. Additionally, it adds an economic benefit over the Peak Search Technique in that more crashes are identified in smaller segment lengths. A comparison of crash identification based on sliding window and peak search methods two is shown below.
The top illustration is an example of the Peak Search Method and the lower illustration is an example of the Sliding Window Method. The advantage of the Sliding Window Method over the Peak Search Method is clear from this illustration in that the moving window positions identify more crashes for the same window length.
However, the Sliding Window Method takes more computational time since the number of window positions to aggregate performance measures may increase drastically depending on user specified incremental length.
