Thomas Lumley • 10/14/2022

Code archaeology: polynomial distributed lags

The article details polynomial distributed lag models, a regularization technique for time-series data where lagged predictors are highly correlated. It covers the mathematical basis, the author's experience updating 20-year-old R code for modern use, and compares the approach to modern methods like penalized splines, while discussing implementation challenges in R.

0 comments

#R #Statistical Modeling #Time Series