ICOMP 2024
October 10–12 / Innopolis, Russia
The 1st International Conference on Computational Optimization (ICOMP 2024) will be held in Innopolis, Russian Federation October 10–12, 2024, and is planned to be an in person conference with virtual elements.
We invite submissions of papers on all topics related to optimization. All papers will be reviewed in a double-blind process and accepted papers will be presented at the conference. Papers need to be prepared and submitted as a single file: 9 pages as main paper, with unlimited pages for references and appendix. There will be no separate deadline for the submission of supplementary material. In addition, we require that, barring exceptional circumstances (such as visa problems) upon the acceptance of their papers, at least one of the authors must attend the conference, in person.
Featured proceedings of the ICOMP2024 might be published in special issues of the following journals:
- Journal of Optimization Theory and Applications – Q1 SJR
- Optimization Methods and Software – Q1 SJR
- Computational Mathematics and Mathematical Physics – Q2 SJR
ASCOMP 2024 is a satellite event of ICOMP 2024. The School provides an in-depth introduction to different areas of machine learning and optimization. Its program will comprise four mini-courses from leading AI researchers. Participants will have the opportunity to present their work in a poster session.
The School is open to students and young researchers interested in AI.
- Efficient training and inference methods
- IoT
- Edge Computing
- Novel architectures
- Machine Learning with new compute paradigms (optical, quantum, etc.)
- Data and model compression
- Optimization for Neural Networks
- Dimensionality Reduction
- Feature Selection
- Large Scale, Parallel, and Distributed Algorithms
- Model Selection
- Online Learning
- Software Packages
- Computational Learning Theory
- Optimization for Representation Learning
- Implementation issues
- Efficient and scalable vision
- Transfer/ low-shot/ continual/ long-tail/ learning
- Infrastructure (e.g., libraries, improved implementation and scalability, distributed solutions)
- Applications in any field
As noted above, this year, ICOMP will use a single paper submission deadline with a single review cycle, as follows:
- May 1 – Submissions open
- Aug 4 – Full paper submission deadline
Style Files and an Example Paper. Submitted papers that do not conform to these policies will be rejected without review. Authors are kindly asked to make their submissions as accessible as possible for everyone including people with disabilities and sensory or neurological differences.
Double submission is available.
In no circumstances will extensions be given
Time Zone: Anywhere on Earth
Aug 31 – Registration Cancellation Deadline
Aug 4 Aug 7 – Abstract Submission Deadline
Aug 4 Aug 7 – Full Paper Submission Deadline
Aug 19 Aug 25 – Review release to authors
Sep 2 Sep 5 – Author rebuttal period ends
Sep 10 Sep 12 – Paper Decision notification
Sep 16 Oct 4 – Camera-ready version deadline
Keynote speakers
Conference Hall, 2nd floor, Technopark named after A.S.Popov (7 Universitetskaya St.)
The broadcast will be in the official group of the University
Breakfast
9:00 - 9:30
Transfer to Popov Technopark
9:30 - 10:00
Registration
10:00 - 10:15
Opening
10:15 - 11:00
Peter Richtarik
The First Optimal Parallel SGD (in the Presence of Data, Compute and Communication Heterogeneity)
11:00 - 11:45
Francesco Orabona
Understanding Adam and AdamW through Proximal Updates, Scale-Freeness, and Relaxed Smoothness
11:45 - 12:30
Martin Takac
Elevating Federated Learning: A Shift Towards Second-Order Optimization Methods
12:30 - 13:00
Coffee break
13:00 - 13:30
Alexander Beznosikov
Navigating Byzantine Attacks in Distributed Learning: a Trial Function Methodology
13:30 - 14:00
Alexey Naumov
Stochastic approximation: old and new
14:00 - 15:10
Lunch
Innopolis University
15:10 - 15:40
Andrey Kuznetsov
Multimodal architectures and context compression methods (online)
15:40 - 16:30
Boris Mordukhovich
Convergence of Descent Methods under Polyak-Kurdyka-Lojasiewicz Properties (online)
16:30 - 17:00
Vladimir Spokoiny
Inference for nonlinear inverse problems
17:00 - 17:45
Gesualdo Scutari
Bringing Statistical Thinking in Decentralized Optimization. Vignettes from High-dimensional Statistics over Networks (online)
17:45 - 18:30
Angelia Nedich
Characterizing Trust and Resilience in Distributed Optimization for Cyberphysical Systems (online)
18:30 - 19:00
Coffee break
19:00 - 20:30
Selected reports
Breakfast
9:00-9:20
Transfer to Popov Technopark
9:30 - 10:00
Alexander Gasnikov
Survey of modern results in gradient-free convex optimization
10:00 - 10:30
Roland Hildebrand
Geometric view-points on barrier functions as driver of recent and future progress
10:30 - 11:00
Evgeny Burnaev
Adversarial Schrödinger Bridge Matching for Fast Generative Modeling
11:00 - 11:30
Ivan Oseledets
Effective tensor and matrix methods in deep learning
11:30 - 12:00
Coffee break
12:00 - 12:30
Vladimir Temlyakov
Some remarks on optimization on Banach spaces
12:30 - 13:00
Dmitry Yarotsky
SGD with memory: fundamental properties and stochastic acceleration
13:00 - 13:30
Denis Dimitrov
Diffusion (and more) text-to-video generation methods
13:30 - 14:00
Andrey Vorobyev
Optimization problems for analog devices linearization in wireless communication systems
14:00 - 14:30
Evgeny Tyrtyshnikov
Optimization and tensor decompositions
14:30 - 15:40
Lunch
15:40 - 19:30
Poster session
Breakfast
9:00 - 10:00
Check-out
10:00 - 15:00
Excursion
Registration
Participation format:
- Online listener
- Offline participation with report presentation
- Offline participation without report presentation
Registration fee - $100,
PhD students - $50, students - $15.
Registration fee is paid via Timepad (Russian card transfer) / via bank details (international transfer) only upon receipt of confirmation of acceptance of the article for publication.
By submitting this form you agree to the privacy policy
Payment acceptance partner - IP Laboratory LLC, 7729474160
Organizers
Program Committee
Organizing Committee
-
Darina VenediktovaInnopolis University
Partners
