- Machine-learning-nederland

NeurIPS 2020 Debriefing - Update and Reminder
by Jack Mayo 24 Feb '21

24 Feb '21

Dear Dutch machine learners, This year's national NeurIPS debriefing will take place on: Friday March 5, 2021 14h00-17h00 Via the Following Zoom link: https://uva-live.zoom.us/j/89293881695 PhD students and/or senior researchers will briefly present the paper they found the most interesting at NeurIPS 2020. If you are interested in NeurIPS, please consider being one of the presenters. There are still slots available. The format is to have 3 hours of short talks (15 or 20 minutes, in English). It is not required to have attended NeurIPS, and you would usually not present your own paper. Talks can be informal, in a friendly atmosphere, so this is an ideal opportunity for PhD students to gain experience in giving presentations. The goal is to have presentations from multiple universities, and foster diologue between subdiciplines of the Dutch machine learning community. If you are interested in giving a talk or if you have any questions, feel free to contact me at j [dot] j [dot] mayo [at] uva [dot] nl. Preliminary list of speakers, and topics: Mustafa Celikok, TU Delft - A Unifying View of Optimism in Episodic Reinforcement Learning by Gergely Neu, Ciara Pike-Burne Alexander Mey, TU Delft - Towards Minimax Optimal Reinforcement Learning in Factored Markov Decision Processes by Yi Tian, Jian Qian, Suvrit Sra Jack Mayo, University of Amsterdam - Optimal Algorithms for Stochastic Multi-Armed Bandits with Heavy Tailed Rewards by Kyungjae Lee, Hongjun Yang, Sungbin Lim, Songhwai Oh For up to date information and an overview of past meetings, see www.timvanerven.nl/neurips-debriefing/<https://eur04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.timvan…> Best regards, Jack Mayo and Tim van Erven

1 0

Two funded PhD positions on AI for Biosciences (scheduling or reinforcement learning) at TU Delft, Algorithmics group - deadline 15 March
by Matthijs Spaan 22 Feb '21

22 Feb '21

TU Delft (Netherlands) offers two fully-funded PhD positions as part of the Artificial Intelligence Lab for Biosciences. The first PhD candidate will develop novel approaches for optimal batch scheduling in plants. The second PhD candidate will develop novel reinforcement learning algorithms for self-driving labs. These PhD positions are part of the Artificial Intelligence Lab for Biosciences, the AI4B.io Lab (www.ai4b.io), that TU Delft and Royal DSM recently established. Additionally, 3 PhD positions on machine learning are available within the lab. It is the first of its kind in Europe to apply artificial intelligence to full- scale biomanufacturing, from microbial strain development to process optimization and scheduling. The AI4B.io Lab is part of the Dutch National Innovation Center for AI (ICAI) and starts with five synergistic research projects. You will contribute to DSM’s challenges regarding developing bio-based products and optimizing industrial-scale biobased processes. You will have access to real industrial R&D and/or factory data, work in an industrial as well as an academic environment, and have the opportunity to develop entrepreneurial skills as the AI4B.io Lab collaborates with the Biotech Campus Delft and Planet b.io; an excellent learning and research environment. Further details and application forms can be found via the following links. Smart plant scheduling: https://www.tudelft.nl/over-tu-delft/werken-bij-tu-delft/vacatures/details?… Reinforcement learning for a self-driving lab: https://www.tudelft.nl/over-tu-delft/werken-bij-tu-delft/vacatures/details?… Application deadline for full consideration: 15 March 2021 Informal enquiries: Mathijs de Weerdt (m.m.deweerdt(a)tudelft.nl) for the first position, Matthijs Spaan (m.t.j.spaan(a)tudelft.nl) for the second position.

1 0

Fwd: Craig Boutilier @ Challenges and Opportunities in Multiagent RL
by Frans Oliehoek 10 Feb '21

10 Feb '21

FYI. Please forward to possibly interested people. -------- Forwarded Message -------- Subject: Craig Boutilier @ Challenges and Opportunities in Multiagent RL Date: Thu, 4 Feb 2021 14:08:43 +0100 From: Frans Oliehoek <fa.oliehoek(a)gmail.com> To: Frans Oliehoek - EWI <F.A.Oliehoek(a)tudelft.nl> CC: c.amato(a)northeastern.edu, garnelo(a)google.com, f.a.oliehoek(a)tudelft.nl, somidshafiei(a)google.com, karltuyls(a)google.com Dear all, After a fantastic inaugural presentation by Michael Bowling, we are excited to announce the next speaker in our virtual seminar series on the Challenges and Opportunities for Multiagent Reinforcement Learning (COMARL): Speaker: Craig Boutilier, Google Research Title: Maximizing User Social Welfare in Recommender Ecosystems (abstract and bio can be found below) Date: Thursday February 11th, 2021 Time: 17:00 CET / 16:00 UTC / 08:00 PST Location: via google meet or youtube For detailed instructions on how to join, please see here: https://sites.google.com/view/comarl-seminars/how-to-attend For additional information, please see our: * Website <https://sites.google.com/view/comarl-seminars>(includes schedule, instructions on how to join, etc.) * Twitter account (for speaker announcements and more!):@ComarlSeminars <https://twitter.com/ComarlSeminars> * Google Groups (to receive invitations): comarlseminars(a)googlegroups.com <mailto:comarlseminars@googlegroups.com> We look forward to seeing you there! Best regards from the organizers, Chris Amato (Northeastern University), Marta Garnelo (DeepMind), Frans Oliehoek (TU Delft), Shayegan Omidshafiei (DeepMind), Karl Tuyls (DeepMind) Speaker: Craig Boutilier Google Research, Mountain View, CA, USA Title: Maximizing User Social Welfare in Recommender Ecosystems Abstract: An important goal for recommender systems is to make recommendations that maximize some form of user utility over (ideally, extended periods of) time. While reinforcement learning has started to find limited application in recommendation settings, for the most part, practical recommender systems remain "myopic" (i.e., focused on immediate user responses). Moreover, they are "local" in the sense that they rarely consider the impact that a recommendation made to one user may have on the ability to serve other users. These latter "ecosystem effects" play a critical role in optimizing long-term user utility. In this talk, I describe some recent work we have been doing to optimize user utility and social welfare using reinforcement learning and equilibrium modeling of the recommender ecosystem; draw connections between these models and notions such as fairness and incentive design; and outline some future challenges for the community. Bio: Craig Boutilier is a Principal Scientist at Google. He received his Ph.D. in Computer Science from U. Toronto (1992), and has held positions at U. British Columbia and U. Toronto (where he served as Chair of the Dept. of Computer Science). He co-founded Granata Decision Systems, served as a technical advisor for CombineNet, Inc., and has held consulting/visiting professor appointments at Stanford, Brown, CMU and Paris-Dauphine. Boutilier's current research focuses on various aspects of decision making under uncertainty, including: recommender systems; user modeling; MDPs, reinforcement learning and bandits; preference modeling and elicitation; mechanism design, game theory and multi-agent decision processes; and related areas. Past research has also dealt with: knowledge representation, belief revision, default reasoning and modal logic; probabilistic reasoning and graphical models; multi-agent systems; and social choice. Boutilier served as Program Chair for IJCAI-09 and UAI-2000, and as Editor-in-Chief of the Journal of AI Research (JAIR). He is a Fellow of the Royal Society of Canada (FRSC), the Association for Computing Machinery (ACM) and the Association for the Advancement of Artificial Intelligence (AAAI). He also received the 2018 ACM/SIGAI Autonomous Agents Research Award.

1 0

Two funded PhD positions on Epistemic AI (constraint programming or reinforcement learning) at TU Delft - deadline 28 February
by Matthijs Spaan 04 Feb '21

04 Feb '21

TU Delft (Netherlands) offers two fully-funded PhD positions as part of the EU FET Open project Epistemic AI. The first PhD candidate will develop novel approaches for combinatorial optimisation under epistemic uncertainty. The second PhD candidate will develop novel reinforcement learning algorithms that aim for robust and safe behaviour in partially-known environments. The project goal is to create a new paradigm for next-generation AI providing worst-case guarantees on its predictions thanks to a proper modelling of real-world uncertainties. Further details and application form: https://www.tudelft.nl/over-tu-delft/werken-bij-tu-delft/vacatures/details?… Application deadline for full consideration: 28 February 2021 Informal enquiries: Neil Yorke-Smith (n.yorke-smith(a)tudelft.nl) for the first position, Matthijs Spaan (m.t.j.spaan(a)tudelft.nl) for the second position.

1 0

[jobs] PostDocs@TU Delft - Interactive Robot Learning
by Jens Kober - 3ME 04 Feb '21

04 Feb '21

Topic: 'Interactive Robot Learning', see job description below Location: TU Delft, The Netherlands Duration: 24 months Application deadline: March 15th 2021 Starting date: at the latest September 2021 Salary: EUR 3.491 - EUR 4.402 For more details see https://www.tudelft.nl/over-tu-delft/werken-bij-tu-delft/vacatures/details?… Applications need to be submitted via the website above. For more information about this vacancy, please contact Jens Kober, Associate Professor, email: mailto:J.Kober@tudelft.nl *Job description* Programming and re-programming robots is extremely time-consuming and expensive, which presents a major bottleneck for new industrial, agricultural, care, and household robot applications. The goal of this project is to enable robots to learn how to perform manipulation tasks from few human demonstrations, based on novel interactive machine learning techniques. Robot learning will no longer rely on initial demonstrations only, but it will effectively use additional user feedback to continuously optimize the task performance. It will enable the user to directly perceive and correct undesirable behavior and to quickly guide the robot toward the target behavior. You will explore one or several aspects of interactive robot learning: learning force-interaction skills with user inputs, requesting additional advice, interactive imitation and reinforcement learning for sequences, interactive inverse reinforcement learning, and evaluating how humans prefer to teach robots. You will evaluate the developed approaches with generic real-world robotic force-interaction tasks related to handling and (dis)assembly. You will demonstrate the potential of the newly developed teaching framework with challenging bi-manual tasks and a final user study evaluating how well novice human operators can teach novel tasks to a robot. The Postdoc positions are in the context of the project "Teaching Robots Interactively" (TERI), funded by the European Research Council as ERC Starting Grant. *Requirements* You have a PhD degree in systems and control, robotics, applied mathematics, artificial intelligence, machine learning, or a related subject. You must have strong analytical skills and must be able to work at the intersection of several research domains. Experience with real robot applications, bi-manual robots, interactive learning, and/or user studies is a plus. You must have demonstrated ability to conduct high-quality research according to international standards, as demonstrated by publications in international, high-quality journals. A very good command of the English language is required, as well as excellent communication skills.

1 0

[jobs] PhD position in causality-inspired ML at University of Amsterdam
by Sara Magliacane 29 Jan '21

29 Jan '21

*Job offer:* fully-paid 4-year PhD position under the Collective Labour Agreement of Dutch Universities (solid pension scheme, a maximum of 41 days of annual leave based on a 40 hours work week, paid sick leave, maternity and paternity leave) *Research group*: INtelligent Data Engineering Lab (indelab.org), Informatics Institute, University of Amsterdam (uva.nl/en) *Deadline*: April 15th 2021 *Project*: Causality-inspired ML Are you interested in applying ideas from causal inference in machine learning research? The INtelligent Data Engineering Lab at the University of Amsterdam is seeking a PhD candidate in causality-inspired machine learning (ML), i.e. the application of ideas from causality to different areas of ML, under the supervision of Dr Sara Magliacane ( smaglia.wordpress.com). In particular, we are looking for PhD students that are interested in exploring the connections between causal inference, transfer learning and active learning. *Application link and details*: https://www.uva.nl/shared-content/uva/en/vacancies/2021/01/21-058-phd-candi…

1 0

Women in Data Science Conference 2021
by Info-ids (IDS) 29 Jan '21

29 Jan '21

Institute of Data Science at Maastricht University Netherlands organizes the second edition of WiDS Maastricht in conjunction to WiDS Netherlands. The two events are co-joint and will take place online on 8 and 9 March 2021! The Women in Data Science (WiDS) conference is a technical conference that provides an opportunity to hear about the latest data science related research and to connect with others in the field. All genders are invited to attend these WiDS events. On Day 1 (Monday 8 March) WiDS Netherlands partakes in the first-ever 24-hour virtual WiDS Worldwide Initiative (hyperlink: https://www.widsconference.org) The speakers will highlight data science works that impact society. On Day 2 (Tuesday 9 March) WiDS Maastricht will follow. The speakers and the audience will focus on the impact of data science on society and on responsible data science. WiDS started as a conference at Stanford in November 2015. In 2020 the WiDS Worldwide conference reached over 100,000 people across the globe and organised 150+ WiDS regional events in 50+ countries. For more information and registrations please check here https://www.maastrichtuniversity.nl/events/wids-conference-2021

1 0

Postdoctoral position at the University of Michigan
by Pires da Silva Castro, R.M. 26 Jan '21

26 Jan '21

Posting on behalf of Laura Balzano (University of Michigan) ---------------------- Begin forwarded message: From: Laura Balzano <girasole(a)umich.edu<mailto:girasole@umich.edu>> Subject: Fwd: Postdoctoral position at UM Date: 25 January 2021 at 23:34:46 CET To: undisclosed-recipients:; Dear colleagues, I have a postdoc opportunity at the University of Michigan to begin in spring 2021, working with me and also with the hope that the postdoc also establishes a co-mentor at UM. Please share with your colleagues and students and have them send their CV and research statement to me at <girasole(a)umich.edu<mailto:girasole@umich.edu>> with the subject "Joining the Balzano lab -- postdoc 2021" if they are interested. I am seeking a postdoc who is interested in optimization methods for novel machine learning applications. Any topic under the umbrella of low-dimensional modeling would be appropriate for this postdoctoral position. I am particularly interested in optimization for constrained SVD, nonlinear low-dimensional models, nonlinear or non-independent measurement models, joint dimensionality reduction and clustering, sketching within optimization algorithms, and time-varying low-dimensional models. Applications that motivate my work recently include power systems monitoring, medical and computational imaging, preference learning and learning to rank, control systems identification, and environmental monitoring. The position will be one year renewable for three (two is the most likely length). I would be thrilled to receive applications from people in underrepresented groups in machine learning, such as women and BIPOC candidates. This postdoc has no requirements other than research. As I am keenly aware of extra service and outreach workload burden for these groups, I do my best to advise those in my group on these matters and help them protect their time. Thank you, Laura Balzano Associate Professor of Electrical Engineering and Computer Science http://web.eecs.umich.edu/~girasole/<https://eur02.safelinks.protection.outlook.com/?url=http:%2F%2Fweb.eecs.umi…>

1 0

NeurIPS 2020 - Debriefing Invitation
by Jack Mayo 19 Jan '21

19 Jan '21

Dear Dutch machine learners, There will be a national NeurIPS debriefing again, where PhD students and/or senior researchers will briefly present the paper they found the most interesting at NeurIPS 2020. If you are interested in NeurIPS, please consider being one of the presenters. The format is to have 2.5 hours of short talks (15 or 20 minutes, in English). It is not required to have attended NeurIPS, and you would usually not present your own paper. Talks can be informal, in a friendly atmosphere, so this is an ideal opportunity for PhD students to gain experience in giving presentations. An exact date and schedule will be announced later, but we are currently thinking to set the date in late February. If you are interested in giving a talk or if you have any questions, please let one of us know. For up to date information and an overview of past meetings, see www.timvanerven.nl/neurips-debriefing/<https://eur04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.timvan…> Best regards, Jack J. Mayo (j.j.mayo [at] uva [dot] nl) and Tim van Erven (tim [at] timvanerven [dot] nl)

1 0

Reinforcement Learning challenge (Turing Institute, Jan 18-25)
by Alessandro Zocca 07 Jan '21

07 Jan '21

Dear colleagues, Would you or your students like to gain hands-on experience of Reinforcement Learning? Take part in testing RangL, a brand new open-source competition platform to accelerate progress in data-driven control problems, especially in the context of energy and transportation systems. Reinforcement Learning challenge 18-25 Jan 2021 The Alan Turing Institute • Free entry • Individuals and teams (two or more people) can apply • Three levels of difficulty available • All successful invitees supported with ‘Ask away!’ channel on Slack • Certificate of participation included Contact: rangl(a)turing.ac.uk before 8 January 2021. Participation is by invitation only so early application is advised. Check out the project web page at https://www.turing.ac.uk/research/research-projects/ai-control-problems ================================= Alessandro Zocca (he/his) Assistant Professor Vrije Universiteit Amsterdam Department of Mathematics https://sites.google.com/site/zoccaale/

1 0

2024

2023

2022

2021

2020

Machine-learning-nederland