The Partial Least Squares Spline Model for Public Health Surveillance Data

Algorithm 1

Partial least squares spline (PLS-spline) model.

1: function PLS model where is the covariate matrix, is the time, is the event, and is the number of components.

2:   loading weights

3:   normalized loading weights

4:   score vector

5:  X-loadings

6:   -loadings     ⊳ repeat the above steps until

7:  for to do

8:       ⊳ Royston and Parmer (RP) restricted cubic spline model on PLSR components.