Conditional Super Learner
This addresses model selection challenges in machine learning, particularly for hierarchical problems, but appears incremental as it builds on existing stacking and meta-learning concepts.
The paper tackles the problem of selecting the best model from a library based on covariates by proposing the Conditional Super Learner (CSL) algorithm, which merges cross-validation with meta learning and is shown to converge faster than O_p(n^{-1/4}) while offering empirical evidence as a substitute for stacking or hierarchical analysis.
In this article we consider the Conditional Super Learner (CSL), an algorithm which selects the best model candidate from a library conditional on the covariates. The CSL expands the idea of using cross-validation to select the best model and merges it with meta learning. Here we propose a specific algorithm that finds a local minimum to the problem posed, proof that it converges at a rate faster than $O_p(n^{-1/4})$ and offers extensive empirical evidence that it is an excellent candidate to substitute stacking or for the analysis of Hierarchical problems.