Call for papers: 7th ICML Workshop on Automated Machine Learning (AutoML 2020)
Frank Hutter
Machine learning has achieved considerable successes in recent years, but this success often relies on human experts, who construct appropriate features, design learning architectures, set their hyperparameters, and develop new learning algorithms. Driven by the demand for off-the-shelf machine learning methods from an ever-growing community, the research area of AutoML targets the progressive automation of machine learning aiming to make effective methods available to everyone. Hence, the workshop targets a broad audience ranging from core machine learning researchers in different fields of ML connected to AutoML, such as neural architecture search, hyperparameter optimization, meta-learning, and learning to learn, to domain experts aiming to apply machine learning to new types of problems.
We invite submissions on the topics of:
- Model selection, hyper-parameter optimization, and model search
- Neural architecture search
- Meta-learning and transfer learning
- Bayesian optimization for AutoML
- Evolutionary algorithms for AutoML
- Multi-fidelity optimization
- Predictive models of performance
- Automatic feature extraction / construction
- Automatic data cleaning
- Automatic generation of workflows / workflow reuse
- Automatic problem "ingestion" (from raw data and miscellaneous formats)
- Automatic feature transformation to match algorithm requirements
- Automatic acquisition of new data (active learning, experimental design)
- Automatic report generation (providing insight on automatic data analysis)
- Automatic selection of evaluation metrics / validation procedures
- Automatic selection of algorithms under time/space/power constraints
- Automatic construction of fair and unbiased machine learning models
- Automation of semi-supervised and unsupervised machine learning
- Demos of existing AutoML systems
- Robustness of AutoML systems (w.r.t. Randomized algorithms, data, hardware etc.)
- Human-in-the-loop approaches for AutoML
- Learning to learn new algorithms and strategies
- Hyperparameter agnostic algorithms
We welcome submissions of up to 6 pages in JMLR Workshop and
Proceedings format (plus 10 pages for references and appendix).
All accepted papers will be presented as posters. We will invite
the 2-3 best papers for an oral plenary presentation. We will provide PDFs of accepted
papers on http://icml2020.automl.org/, but there will be no archival proceedings. For submission details
please see the submission page.
Due to the Covid-19 pandemic, ICML will be an entirely virtual conference this year (https://icml.cc/Conferences/2020/VirtualICML), and the AutoML workshop will follow suit. We are in contact with the ICML organizers about how this will pan out exactly and will inform you as soon as possible.
We provide up-to-date information on http://icml2020.automl.org
Keynote Speakers
- Alex Smola (Director, Amazon Web Services)
- Mihaela van der Schaar (Professor of Machine Learning, Artificial Intelligence and Medicine at the University of Cambridge)
- Neil Lawrence (DeepMind Professor of Machine Learning at the University of Cambridge and visiting Professor at the University of Sheffield)
Organization
Location
The 7th ICML workshop on AutoML will be co-located with the 37th
International Conference on Machine Learning (ICML 2020) and will
take place on July 17th or 18th.
Important Dates
Deadline: April 23rd (Anywhere on earth)Notification: May 25th (Anywhere on earth)
On behalf of the
organizing committee,
Frank Hutter
