SAM provides sparse additive models for high-dimensional prediction tasks (regression and classification). It uses spline basis expansion and efficient optimization routines to compute full regularization paths.
Tuo Zhao, Xingguo Li, Haoming Jiang, Han Liu, and Kathryn Roeder
Maintainer: Tuo Zhao <tourzhao@gatech.edu>
The package exposes four model families:
samQL: quadratic-loss sparse additive regression.
samLL: logistic-loss sparse additive classification.
samHL: hinge-loss sparse additive classification.
samEL: Poisson-loss sparse additive regression.
All models share a common spline representation and return regularization paths, allowing model selection after one fit.
P. Ravikumar, J. Lafferty, H.Liu and L. Wasserman. "Sparse Additive Models", Journal of Royal Statistical Society: Series B, 2009.
T. Zhao and H.Liu. "Sparse Additive Machine", International Conference on Artificial Intelligence and Statistics, 2012.
samQL,samHL,samLL,samEL