Pipeline Combinators for Gradual AutoML - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2021

Pipeline Combinators for Gradual AutoML

Guillaume Baudart
  • Function : Author
  • PersonId : 1119186
Martin Hirzel
  • Function : Author
  • PersonId : 1119187
Kiran Kate
  • Function : Author
  • PersonId : 1119188
Avraham Shinnar
  • Function : Author
  • PersonId : 1119189
Jason Tsay
  • Function : Author
  • PersonId : 1119190

Abstract

Automated machine learning (AutoML) can make data scientists more productive. But if machine learning is totally automated, that leaves no room for data scientists to apply their intuition. Hence, data scientists often prefer not total but gradual automation, where they control certain choices and AutoML explores the rest. Unfortunately, gradual AutoML is cumbersome with state-of-the-art tools, requiring large non-compositional code changes. More concise compositional code can be achieved with combinators, a powerful concept from functional programming. This paper introduces a small set of orthogonal combinators for composing machinelearning operators into pipelines. It describes a translation scheme from pipelines and associated hyperparameter schemas to search spaces for AutoML optimizers. On that foundation, this paper presents Lale, an open-source sklearn-compatible AutoML library, and evaluates it with a user study.
Fichier principal
Vignette du fichier
neurips21.pdf (546.08 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03464012 , version 1 (02-12-2021)

Identifiers

  • HAL Id : hal-03464012 , version 1

Cite

Guillaume Baudart, Martin Hirzel, Kiran Kate, Parikshit Ram, Avraham Shinnar, et al.. Pipeline Combinators for Gradual AutoML. NeurIPS 2021 - Thirty-fifth Conference on Neural Information Processing Systems, Dec 2021, Virtual, France. ⟨hal-03464012⟩
40 View
78 Download

Share

Gmail Mastodon Facebook X LinkedIn More