Understanding Order Tracking Algorithm Limitations 2009-01-2157
An in depth discussion is presented as to why some order tracking datasets are difficult to analyze with many if not all order tracking methods. If a solid understanding of a dataset's properties is present, oftentimes it is very clear as to why different order tracking methods behave the way they do. Several analytical datasets will be presented along with comparisons of results achieved using different order tracking methods. In conjunction with these datasets, the algorithm performance of nearly all current order tracking methods will be characterized with respect to the different properties a dataset may possess. It is hoped that through the data, concepts, and examples presented in the paper that the reader will develop a very intuitive sense as to how to approach order tracking analysis on both simple and complex or difficult datasets.