Lessons learned from the AutoML challenge

Lisheng Sun-Hosoya (1) , Isabelle Guyon (2, 1) , Michèle Sebag (1)

1 TAU - TAckling the Underspecified
2 Chalearn

We give a brief account of the main findings of our post-hoc analysis of the first AutoML challenge (2015-2016). This competition, which took place in 2015-2016 challenged the participants to submit code that solve classification and regression problems from fixed-length feature representations, without any human intervention. This paper is a digest of a book chapter to be published in the Springer Series on Challenges in Machine Learning [Geaar]. All datasets, code of the winners, and challenge results are found at: http://automl.chalearn.org.

Mots clés: Automatic ML; Meta learning; Transfer learning; Bayesian optimization

Domaines: Informatique [cs]; Apprentissage [cs.LG]; Intelligence artificielle [cs.AI]

Lisheng Sun-Hosoya, Isabelle Guyon, Michèle Sebag. Lessons learned from the AutoML challenge. Conférence sur l'Apprentissage Automatique 2018, Jun 2018, Rouen, France. ⟨hal-01811454⟩