Estimating national and state sales for a class of products required models using biased data.
A machine/statistical-learning algorithm, “Elastic Net,” was used to optimize the choice of a model. This would extrapolate sales (from a biased sample of convenience) to the entire country and aggregate them down to sub-state-geographic levels. An extension of the method creates forecasts to future periods.
The predictive power and scope of applying an existing model was increased. The new one creates accurate sales estimates, along with statistical indications of their precision (in the form of standard errors). The estimates are consistent over time, permitting useful year-over-year growth comparisons and market share estimates.