The goal of transfer learning is to improve the learning of a new target concept given knowledge of related source concept(s). We introduce the first boosting-based algorithms for transfer learning that apply to regression tasks. First, we describe two existing clas- sification transfer algorithms...
Boosting Classifiers with Tightened L0-Relaxation Penalties (ICML 2010) Noam Goldberg, Jonathan Eckstein [Paper] Boosting for Regression Transfer (ICML 2010) David Pardoe, Peter Stone [Paper] [Code] Boosted Backpropagation Learning for Training Deep Modular Networks (ICML 2010) Alexander Grubb, J...
where no distributional assumption on the effect sizes is made. Although limited from the computational perspective due to the extremely high-dimensional data in GWAS, high-dimensional linear regression is a natural model for GWAS in modelling the whole-genome level contributions of genetic variation...