Our quest for robust time series forecasting at scale
The Unofficial Google Data Science Blog
APRIL 17, 2017
Quantification of forecast uncertainty via simulation-based prediction intervals. Such a model risks conflating important aspects, notably the growth trend, with other less critical aspects. In other words, there is an asymmetry of risk-reward when there exists the possibility of misspecifying the weights in $X_C$.
Let's personalize your content