Effects of parameter selection on forecast accuracy and execution time in nonparametric regression

Recent research has shown nonparametric regression to hold high potential to accurately forecast short-term traffic flow. Nonparametric regression is a forecasting technique based on nearest neighbor searching in which forecasts are derived from past observations that are similar to the current conditions. However, many practical, fundamental questions remain, such as how to reduce execution times while addressing concerns about acceptable designs and implementations of the nonparametric regression algorithm. The results presented indicate that advanced data structures can significantly reduce the execution time of nearest neighbor nonparametric regression. Further reductions in execution time may be achieved through the use of approximate nearest neighbors, but at the expense of forecast accuracy.