Trevor Hastie

Overview

Trevor Hastie is a statistician (Stanford University), one of the most influential figures in statistical learning — co-author of The Elements of Statistical Learning and An Introduction to Statistical Learning, and a developer of generalized additive models, LARS, and the glmnet regularization-path software.

Key Contributions / Features

• Elastic net (Zou-Hastie 2005): the $\ell_1+\ell_2$ penalized regression with the grouping effect and $p\gg n$ capability.
• LARS (Efron-Hastie-Johnstone-Tibshirani 2004) and glmnet (Friedman-Hastie-Tibshirani 2010) coordinate-descent solvers for lasso/elastic-net paths.
• Generalized additive models (Hastie-Tibshirani 1990); foundational statistical-learning textbooks.

Related

• Lasso
• Coordinate Descent
• Zou-Hastie (2005) Regularization and Variable Selection via the Elastic Net
• Friedman-Hastie-Tibshirani (2010) Regularization Paths for Generalized Linear Models via Coordinate Descent
• Jerome Friedman
• Robert Tibshirani
• Hui Zou
• P-splines