AI Autumn School

Elevate Your AI Expertise with ASCOMP

Gain an unparalleled depth in Machine Learning and Optimization through our intensive program. ASCOMP 2025 features exclusive mini-courses taught by world-leading AI researchers. Network with top-tier students and young researchers, and present your work in our dedicated poster session. ASCOMP is satellite event of ICOMP 2025 hosted by Khalifa University, Abu Dhabi

AI Autumn School

on Computational Optimization

Abu Dhabi

14–16 October 2025

instructors

Volkan Cevher

LIONS, EPFL
Daniel Kuhn

EPFL
"Distributionally Robust Optimization"
Selin Ahipasaoglu

University of Southampton
“Minimum Volume Ellipsoids: Formulations, applications, theory, and methods”
Robert M. Gower

Flatiron Institute
"The Muon Optimizer and Non-Euclidean gradient descent for Neural Networks"
Katya Scheinberg

Georgia Tech
"Introduction to derivative-free and zeroth order optimization"

Alexander Gavrilenko

DBP-Finder: Enhanced Identification of DNA-Binding Proteins Using Fine-Tuned Protein
Language Models

link to poster
Sergey Vandanov

Parallel Clustering Algorithm for the k-medoids Problem in High-dimensional Space for Large-scale Datasets

link to poster
Artyom Matveev

Rejection-based EEG Signal Classification

link to poster
Timur Bavshin

Study of optimization methods in the task of segmentation and defect detection in structural materials

link to poster
Aleksandr Lobanov

Acceleration Exists! Optimization Problems When Oracle Can Only Compare Objective Function Values

link to poster
Aleksandr Maslovskiy

Comparison of Traditional and AI-based Models for IMD2 Cancellation

link to poster
Andrey Sadchikov

Local SGD converges faster for quadratic-like objectives and requires less communication

link to poster
Andrey Veprikov

A Mathematical Model of the Hidden Feedback Loop Effect in Machine Learning Systems

link to poster
Daniil Medyakov

Shuffling Heuristic in Variational Inequalities Establishing New Convergence Guarantees

link to poster
Danila Rodionov

Operator learning for constructing transparent BC for hyperbolic equations

link to poster
Demyan Yarmoshik

Decentralized Optimization with Coupled Constraints

link to poster
Denis Samatov

Automatic segmentation of epicardial fat and quantification of radiomic parameters in cardiac computed tomography

link to poster
Dmitry Bylinkin

Bregman Proximal Method for Efficient Communications under Similarity

link to poster
Elfat Sabitov

Path-based approach for Important Node Exploration

link to poster
Gleb Molodtsov

Shuffling Heuristic in Variational Inequalities Establishing New Convergence Guarantees

link to poster
Ilia Chichkanov

Longitudinal Movement Adaptive Control

link to poster
Irina Maltseva

LLM in RecSys

link to poster
Julia Kudryavtseva

NeRF with semantics for reconstruction of 3D scene and object segmentation by text prompts

link to poster
Kasymkhan Khubiyev

Multimodal approaches in stock price prediction

link to poster
Konstantin Baltsat

Language-dependent Moral Basis of Large Language Models in a Trolley Dilemma

link to poster
Maryna Horbach

Tensor bandits and their applications

link to poster
Meruza Kubentaeva

Primal-Dual Gradient Methods for Searching Network Equilibria in Combined Models with Nested Choice Structure and Capacity Constraints

link to poster
Mikhail Borisov

Estimating cloud base height from all-sky imagery using artificial neural networks

link to poster
Nhat Trung Nguyen

Average-case analysis for distributed consensus algorithms on regular graphs

link to poster
Nikita Kulakov

Interior point methods for mathematical optimization

link to poster
Nikita Pletnev

Application of gradient methods to solving ill-posed problems of mathematical physics

link to poster
Victor Krasnikov

Diffractive neural networks for image processing

link to poster
Yury Svirschevsky

Metropolis-Hastings with Approximate Acceptance Ratio Calculation

link to poster