Data analysts are able to discover relationships that may exist between variables using function-on-function regression methods. Although this statistical method is used in many fields, from business to meteorology, it is typically limited to analyzing relationships between variables as a ‘snapshot’ fixed in time.
Kuang-Yao Lee’s method of nonparametric function-on-function regression incorporates datasets from neighboring points in time to improve the depth of analysis. For example, not only does Lee’s method identify whether hourly temperature affects bicycle rentals, but it also identifies whether there are specific times throughout the day that temperature has the most impact on rentals.
Lee’s method is theoretically solid, can easily be implemented by analysts and is compatible with large data sets.