Schedule

The event is collocated with ECML-PKDD 2017 and will take place on Friday, September 22, 2017.

It will include a tutorial in the morning and a workshop in the afternoon.

Tutorial

9:00 - 10:40 Session 1 (1h40 min)

Introduction

Blackbox hyperparameter optimization and automated machine learning

Algorithm configuration

Meta-Learning for Algorithm Selection

10:40 - 11:00 Coffee break (20 min)

11:00 - 12:40 Session 2 (1h40 min)

Beyond blackbox optimization

AutoML on streaming data

Meta-learning for Automating workflow design

AutoML systems and applications

The big picture: machine learning for automated algorithm design

Presenters: Pavel Brazdil, Joaquin Vanschoren, Frank Hutter, Holger Hoos

Lunch break

12:40 - 14:00 (1h20 min.)

Workshop

14:00 - 15:40 Session 1 (100 min.)

Welcome (5 min.)

Invited Talk by Michele Sebag (30 +10 min.):

Stochastic Gradient Descent: Going as Fast as Possible but not Faster

Poster spotlights 1 (nº 1,3,5,6 in Accepted papers page, 4 min. each)

Poster session 1 (40 min.)

15:40 - 16:00 Coffee break (20 min.)

16:00 - 17:40 Session 2 (100 min.)

G Verbruggen and L de Raedt:

Towards Automated Relational Data Wrangling (20+5 min)

G Contardo, L Denoyer and T Artieres:

A Meta-Learning Approach to One-Step Active-Learning (20+5 min)

Poster spotlights 2 (nº 7,8,9 in Accepted papers page, 4 min. each +

nº 10,11 in Accepted papers page, 2 min. each)

Poster session 2 (35 min.)

17:40-18:40 Session 3 (Extra session) (60 min.)

Poster session 3 (25 min.)

Panel & Commentaries on some papers & research lines (30 min.) -> 18:10

Closing (5 min)