-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
b745178
commit e52cc67
Showing
2 changed files
with
48 additions
and
66 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,56 +1,41 @@ | ||
| --- | ||
| permalink: / | ||
| title: <center>Workshop on Real World Experiment Design and Active Learning at ICML 2020</center> | ||
| title: <center>ICML 2023 Workshop on Adaptive Experimental Design and Active Learning in the Real World</center> | ||
| excerpt: "About" | ||
| sitemap: false | ||
| author_profile: true | ||
| redirect_from: | ||
| - /about/ | ||
| - /about.html | ||
| --- | ||
| #### <center> Virtual workshop, 18 July 2020 @ICML</center> #### | ||
| #### <center> In-Person Workshop, 28-29 July 2023 @ICML</center> #### | ||
| ------ | ||
| <p style="text-align: justify;"> | ||
| This workshop aims to bring together researchers from academia and industry to discuss major challenges, outline recent advances, and highlight future directions pertaining to novel and existing large-scale real-world experiment design and active learning problems. We aim to highlight new and emerging research opportunities for the ML community that arise from the evolving needs to make experiment design and active learning procedures that are theoretically and practically relevant for realistic applications. Progress in this area has the potential to provide immense benefits in using experiment design and active learning algorithms in emerging high impact applications, such as material design, computational biology, algorithm configuration, AutoML, crowdsourcing, citizen science, robotics, and more. </p> | ||
|
|
||
| **Remark:** In light of the COVID-19 situation, the workshop will be held virtually. For more information, please see the ICML conference [website](https://icml.cc/Conferences/2020/Dates). | ||
|
|
||
| Talks | ||
| ------ | ||
| * [[Video](https://slideslive.com/38930821/latent-space-optimization-with-deep-generative-models?ref=account-folder-55847-folders)] **José Miguel Hernández Lobato:** *Latent Space Optimization with Deep Generative Models* | ||
| * [[Video](https://slideslive.com/38930824/designing-bayesianoptimal-experiments-with-stochastic-gradients?ref=account-folder-55847-folders)] **Tom Rainforth:** *Designing Bayesian-Optimal Experiments with Stochastic Gradients* | ||
| * [[Video](https://slideslive.com/38930823/active-learning-of-robot-reward-functions?ref=account-folder-55847-folders)] **Dorsa Sadigh:** *Active Learning of Robot Reward Functions* | ||
| * [[Video](https://slideslive.com/38930819/active-learning-thourgh-physically-embodied-synthesizedfromscratch-queries?ref=account-folder-55847-folders)] **Anca Dragan:** *Active Learning through Physically-embodied, Synthesized-from-"scratch" Queries* | ||
| * [[Video](https://slideslive.com/38930822/uncertainty-quantification-using-martingales-for-misspecified-gaussian-processes?ref=account-folder-55847-folders)] **Aaditya Ramdas:** *Uncertainty Quantification Using Martingales for Misspecified Gaussian Processes* | ||
| * [[Video](https://slideslive.com/38930825/learning-to-manage-inventory?ref=account-folder-55847-folders)] **Shipra Agrawal:** *Learning to Manage Inventory* | ||
| * [Video] **Jennifer Listgarten:** *Machine Learning-based Design (of Proteins, Small Molecules and Beyond)* | ||
| * [Video] **Angela Schoellig:** *Safe and Efficient Active Learning Strategies for Robotics Applications* | ||
| * [[Video](https://slideslive.com/38930826/invited-talk-7?ref=account-folder-55847-folders)] **Pietro Perona:** *Towards Causal Benchmarking of Bias in Face Analysis Algorithms* | ||
| This workshop aims to bring together researchers from academia and industry to discuss major challenges, outline recent advances, and highlight future directions pertaining to novel and existing real-world experimental design and active learning problems. In addition, we aim to highlight new and emerging research opportunities for the machine learning community that arise from the evolving needs to make experimental design and active learning procedures that are theoretically and practically relevant for practical applications. Examples include protein design, causal discovery, drug design, and materials design, to name a few. | ||
| </p> | ||
|
|
||
|
|
||
| Invited Speakers | ||
| ------ | ||
| * **[Shipra Agrawal](http://www.columbia.edu/~sa3305/)** (Columbia University) | ||
| * **[Anca Dragan](https://people.eecs.berkeley.edu/~anca/)** (UC Berkeley) | ||
| * **[Jennifer Listgarten](http://www.jennifer.listgarten.com/)** (UC Berkeley) | ||
| * **[José Miguel Hernández Lobato](https://jmhl.org/)** (University of Cambridge) | ||
| * **[Pietro Perona](http://www.vision.caltech.edu/Perona.html)** (Caltech) | ||
| * **[Tom Rainforth](http://www.robots.ox.ac.uk/~twgr/)** (University of Oxford) | ||
| * **[Aaditya Ramdas](https://www.stat.cmu.edu/~aramdas/)** (Carnegie Mellon University) | ||
| * **[Dorsa Sadigh](https://dorsa.fyi/)** (Stanford University) | ||
| * **[Angela Schoellig](http://www.dynsyslab.org/prof-angela-schoellig/)** (University of Toronto) | ||
| * [Confirmed] **[Mihaela van der Schaar](https://www.vanderschaar-lab.com/)** , Cambridge University. | ||
| * [Confirmed] **[Eytan Bakshy](https://eytan.github.io/)**, Meta Platforms, Inc. | ||
| * [Confirmed] **[Hamsa Bastani](https://hamsabastani.github.io/)**, University of Pennsylvania / Wharton. | ||
| * [Confirmed] **[Sara Magliacane](https://saramagliacane.github.io/)**, University Amsterdam and MIT-IBM Watson. | ||
| * [Confirmed] **[Stefan Bauer](https://www.kth.se/profile/baue)**, KTH Stockholm. | ||
| * [Confirmed] **[Emma Brunskill](https://cs.stanford.edu/people/ebrun/)**, Stanford University. | ||
| * [Invited] **[Nathan Kallus](https://nathankallus.com/)**, Cornell University and Cornell Tech. | ||
|
|
||
|
|
||
| Call for Submissions & Important Dates | ||
| ------ | ||
| Please see the [Call for papers](https://realworldml.github.io/icml2020/cfp/) for submission instructions. | ||
| Please see the [Call for papers](https://realworldml.github.io/icml2022/cfp/) for submission instructions. | ||
| * **Submission deadline:** 3rd June 2022, 11:59 PM (AoE time) | ||
| * **Notification of acceptance:** 13th June 2022, 11:59 PM (AoE time) | ||
|
|
||
| **Best Student Paper Award:** A *best student paper award*, worth **1000 USD**, will be awarded to the best paper selected by a reviewing committee. | ||
|
|
||
| * Submission deadline: 22 June 2020, 11:59 PM (AoE time) | ||
| * [Camera-ready paper](https://realworldml.github.io/icml2020/cfp#camera-ready-instructions) submission deadline: 15 July 2020, 11:59 PM (AoE time) | ||
| * [Lightning talk slides](https://realworldml.github.io/icml2020/cfp/#lightning-talk-instructions) submission deadline: 15 July 2020, 11:59 PM (AoE time) | ||
| * Workshop date: 18th July 2020 | ||
|
|
||
| Organizers | ||
| ------ | ||
| * **[Ilija Bogunovic](https://ilijabogunovic.com)** (ETH Zurich) | ||
| * **[Willie Neiswanger](https://willieneis.github.io/)** (Carnegie Mellon University) | ||
| * **[Yisong Yue](http://www.yisongyue.com/)** (Caltech) | ||
|
|
||
| TODO |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -6,39 +6,36 @@ author_profile: true | |
| sitemap: false | ||
| --- | ||
|
|
||
| <!-- <p style="text-align: justify;"> | ||
| Now is the perfect time for experiment design and active learning methods to make serious headway on a variety of modern challenges and domains. For example, many experiment design tasks involve vast and unstructured search spaces (e.g., the number of potential drug-like molecules is estimated between 10^23 and 10^60), immense amounts of unlabeled data, costly experiments, constraints in the data collection process, the potential for massive parallelism in experimentation, and corrupted or disagreeing measurement outcomes. We are particularly interested in understanding the current issues preventing the broad application of experiment design and active learning methods to these challenges, as well as initial successes.</p> --> | ||
| <!--<p style="text-align: justify;">--> | ||
| <!--</p>--> | ||
| Important Dates | ||
| ---------------- | ||
| * **Submission system opens**: April 20th 2023 11:59 PM (AoE time) [Submission page](https://todo.com){: .btn .btn--warning .btn--large .align-right} | ||
| * **Submission deadline**: June 3rd 2023 11:59 PM (AoE time) | ||
| * **Author notification**: June 13th 2023 11:59 PM (AoE time) | ||
| * **Lightning Talk deadline (spotlight talks)**: TBA | ||
| * **Camera ready date**: TBA | ||
| * **Workshop day**: TBA | ||
|
|
||
| We welcome submissions of 4-6 pages (excluding references) in [JMLR Workshop and Proceedings format](https://www.overleaf.com/latex/templates/template-for-journal-of-machine-learning-research-jmlr-with-jmlr2e-dot-sty/vjcpxhvztrjn). Submissions should be non-anonymous. All accepted papers will be presented as posters (recently published or under-review work is also welcome). There will be no archival proceedings, however, the accepted papers will be made available online on the workshop website. Papers should be submitted [here via EasyChair](https://easychair.org/cfp/realml-icml2020). | ||
| The Call | ||
| --------- | ||
|
|
||
| [Note: the deadline has been **extended by one week**] The work should be submitted by **22nd June 2020, 11:59 PM (Anywhere on Earth)**. | ||
| {: .text-justify} | ||
| Whether in robotics, protein design, or physical sciences, one often faces decisions regarding which data to collect or which experiments to perform. There is thus a pressing need for algorithms and sampling strategies that make intelligent decisions about data collection processes that allow for data-efficient learning. Experimental design and active learning have been major research focuses within machine learning and statistics, aiming to answer both theoretical and algorithmic aspects of efficient data collection processes. The goal of this workshop is to identify missing links that hinder the direct application of these principled research ideas into practically relevant solutions. Progress in this area can provide immense benefits in using experimental design and active learning algorithms in emerging high-impact applications, such as materials design, computational biology, causal discovery, drug design, citizen science, etc. | ||
|
|
||
| Technical topics of interest include (but are not limited to): | ||
| * Large-scale and real-world experiment design (e.g. biological/molecular/drug design, physics, robotics, crowdsourcing, citizen science, algorithms, etc.) | ||
| * Efficient active learning and exploration | ||
| * High-dimensional, scalable Bayesian and bandit optimization (e.g. contextual, multi-task) | ||
| * Sample-efficient interactive learning, hypothesis and A\B testing | ||
| * Corrupted or indirect measurements, multi-fidelity experimentation | ||
| * Incorporating domain-knowledge such as physics | ||
| * Safety and robustness during experimentation and of resulting designs | ||
|
|
||
|
|
||
|
|
||
| ## Camera-Ready Instructions | ||
| Camera-ready papers should be at most 6 pages (excluding references and appendix), using the [customized JMLR template linked here](https://realworldml.github.io/icml2020/_pages/camera_ready_jmlr_tex_template.zip). | ||
| All papers must be sent by **July 15, 2020, 11:59 PM (Anywhere on Earth)** to the following email address: [email protected] | ||
| {: .text-justify} | ||
| We welcome submissions of 4-6 pages (excluding references) in the following (modified) [JMLR Workshop and Proceedings format](https://realworldml.github.io/icml2023/files/author-package.zip). An appendix of any length is allowed after references. Submissions should be non-anonymous. All accepted papers will be presented as posters (recently published or under-review work is also welcome). There will be no archival proceedings, however, the accepted papers will be made available online on the workshop website. Papers should be submitted via [OpenReview](https://todo.com). | ||
|
|
||
| Note that there will be no archival proceedings. However, the camera-ready papers will be made available on the workshop website. | ||
|
|
||
|
|
||
| ## Lightning Talk Instructions | ||
| During the workshop, we will host a round of lightning talks for all accepted papers, where an author from each paper will have the chance to talk about their work live via Zoom, for up to 2 minutes (using slides). Slides must be in PDF format, and sent to organizers in advance, by July 15. Organizers will click through slides while authors speak. Slides will be made available on the workshop website. | ||
| Here are the key details: | ||
| * **Talk date/time:** 5:10 - 6:40 PM UTC on the day of the workshop (July 18, 2020). | ||
| * **Talk details:** Each talk should be under 2 minutes, using any number of slides. Talks should summarize the main ideas and results of the paper. During each talk, organizers will click through slides while authors speak. | ||
| * **Slide format:** Slides should be in PDF format. | ||
| * **Submitting slides:** Slides must be sent by **July 15, 2020 (end of day, Anywhere on Earth)** to the following email address: [email protected] | ||
|
|
||
| Note that we are hosting lightning talks in lieu of a poster session. However, for authors who wish to make and share a poster, we will provide a way for posters to be shared on our website. If you wish to do this, please email us at [email protected]. | ||
| Technical topics of interest include (but are not limited to): | ||
| - Large-scale and real-world experimental design | ||
| (e.g. drug design, physics, robotics, material design, protein design, causal discovery) | ||
| - Efficient active learning and exploration | ||
| - High-dimensional, scalable Bayesian and bandit optimization (e.g. contextual, multi-task) | ||
| - Sample-efficient interactive learning, hypothesis and A/B testing | ||
| - Corrupted or indirect measurements, multi-fidelity experimentation | ||
| - Domain-knowledge integration (e.g. from physics, chemistry, biology, etc.) | ||
| - Safety and robustness during experimentation and of resulting designs | ||
| - Experiment design/active learning in Reinforcement Learning | ||
|
|
||
|
|
||
| Best Paper award | ||
| --------- | ||
| We will be awarding a *best student paper award*, worth **1000 USD**, to the best paper selected by a reviewing committee | ||