Workshop of the International Conference on Machine Learning (ICML) 2017
08:30 AM – 05:30 PM Fri, 11 August @ International Convention Centre, C4.5
TL;DR: 4 pages, in ICML format, submit by June 16 PT.
Deep learning has revolutionized machine learning for many domains and problems. Today, most successful applications of deep learning involve predicting single variables (like univariate regression or multi-class classification). However, many real problems involve highly dependent, structured variables. In such scenarios, it is desired or even necessary to model correlations and dependencies between the multiple input and output variables. Such problems arise in a wide range of domains, from natural language processing, computer vision, computational biology and others.
Some approaches to these problems directly use deep learning concepts, such as those that generate sequences using recurrent neural networks or that output image segmentations through convolutions. Others adapt the concepts from structured output learning. These structured output prediction problems were traditionally handled using linear models and hand-crafted features, with a structured optimization such as inference. It has recently been proposed to combine the representational power of deep neural networks with modeling variable dependence in a structured prediction framework. There are numerous interesting research questions related to modeling and optimization that arise in this problem space.
The workshop will bring together experts in machine learning and application domains whose research focuses on combining deep learning and structured models. Specifically, it will provide an overview of existing approaches from various domains to distill from their success principles that can be more generally applicable. We will also discuss the main challenges arising in this setting and outline potential directions for future progress. The target audience consists of researchers and practitioners in machine learning and application areas.
We invite the submission of short papers no longer than four pages, including references, addressing machine learning research that intersects structured prediction and deep learning, including any of the following topics:
- Deep learning approaches for structured-output problems
- Integration of deep learning with structured-output learning
- End-to-end learning of probabilistic models with non-linear potentials
- Deep learning applications with dependent inputs or outputs
Papers should be formatted according to the ICML template: (http://media.nips.cc/Conferences/ICML2017/icml2017.tgz). Only papers using the above template will be considered. Word templates will not be provided.
Papers should be submitted through easychair at the following address: https://easychair.org/conferences/?conf=1stdeepstructws
Papers will be reviewed for relevance and quality. Accepted papers will be posted online. Authors of high-quality papers will be offered oral presentations at the workshop, and we will award a best-paper and runner-up prize sponsored by Google.
- Abstracts deadline: June 16, 2017 PT
June 2, 2017 PT
- Notification of acceptance: July 1, 2017
June 18, 2017
- Camera-ready deadline: August 1, 2017
- Isabelle Augenstein, University of Copenhagen
- Kai-Wei Chang, University of California at Los Angeles
- Gal Chechik, Bar-Ilan University / Google
- Bert Huang, Virginia Tech
- Andre Martins, Unbabel and Instituto de Telecomunicacoes
- Ofer Meshi, Google
- Yishu Miao, University of Oxford
- Alexander Schwing, University of Illinois Urbana-Champaign