Publications related to Hybrid Decentralized Optimization: Leveraging Both First- and Zeroth-Order Optimizers for Faster Convergence

Related publications (24)

Teaching oxidation states to neural networks

While the accurate description of redox reactions remains a challenge for first-principles calculations, it has been shown that extended Hubbard functionals (DFT+U+V) can provide a reliable approach, mitigating self-interaction errors, in materials with st ...

2025

MARBLE: interpretable representations of neural population dynamics using geometric deep learning

Pierre Vandergheynst

The dynamics of neuron populations commonly evolve on low-dimensional manifolds. Thus, we need methods that learn the dynamical processes over neural manifolds to infer interpretable and consistent latent representations. We introduce a representation lear ...

Nature Publishing Group2025

Foundation neural-networks quantum states as a unified Ansatz for multiple hamiltonians

Luciano Loris Viteritti, Giuseppe Carleo

Foundation models are highly versatile neural-network architectures capable of processing different data types, such as text and images, and generalizing across various tasks like classification and generation. Inspired by this success, we propose Foundati ...

2025

Transferability of Learnt Speech Representations for Decoding Non-Human Vocal Communication

Eklavya Sarkar

Humans and animals both use acoustic signals for vocal communication. The advent of self-supervised learning (SSL) has enabled neural networks to learn robust and general feature representations through the intrinsic acoustic structure of input signals, wi ...

EPFL2025

Intelligent Soft Robotic System for Sensing and Recognition via Triboelectric‐Based Multi‐Sensor Fusion

Xin Chen

Triboelectric nanogenerators (TENGs) serve as a key technology in flexible electronic devices, especially tactile sensors, enabling flexible robotic manipulators to effectively interact with their external environment. Persistent challenges include inadequ ...

Wiley2025

Enhanced Architectures and Optimization Methods for Efficient Language Modeling

Amirkeivan Mohtashami

This thesis investigates methods to construct more capable models more efficiently, focusing on two aspects: improved architectures and optimization. We examine principled architectural modifications that reduce computational costs or introduce features fo ...

EPFL2025

TimeGPT in load forecasting: A large time series model perspective

Fernando Porté Agel, Jiannong Fang, Wenlong Liao

Machine learning models have made significant progress in load forecasting, but their forecast accuracy is limited in cases where historical load data is scarce. Inspired by the outstanding performance of large language models (LLMs) in computer vision and ...

2025

Task-driven neural network models predict neural dynamics of proprioception: Synthetic muscle spindle datasets

Alexander Mathis, Alberto Silvio Chiappa, Alessandro Marin Vargas, Axel Bisi

Here we provide the synthetic spindle datasets of our article "Task-driven neural network models predict neural dynamics of proprioception". It contains the synthetic generated training dataset of simulated muscle spindles during arm passive movements gene ...

Zenodo2024

Topics in statistical physics of high-dimensional machine learning

Hugo Chao Cui

In the past few years, Machine Learning (ML) techniques have ushered in a paradigm shift, allowing the harnessing of ever more abundant sources of data to automate complex tasks. The technical workhorse behind these important breakthroughs arguably lies in ...

EPFL2024

Anatomy-Informed Multimodal Learning for Myocardial Infarction Prediction

Pascal Frossard, Emmanuel Abbé, Ntorina Thanou, Ortal Yona Senouf, Ivan-Daniel Sievering

Goal: In patients with coronary artery disease, the prediction of future cardiac events such as myocardial infarction (MI) remains a major challenge. In this work, we propose a novel anatomy-informed multimodal deep learning framework to predict future MI ...

2024

Recursively Feasible Chance-Constrained Model Predictive Control Under Gaussian Mixture Model Uncertainty

Maryam Kamgarpour, Kai Ren

We present a chance-constrained model predictive control (MPC) framework under Gaussian mixture model (GMM) uncertainty. Specifically, we consider the uncertainty that arises from predicting future behaviors of moving obstacles, which may exhibit multiple ...

Institute of Electrical and Electronics Engineers (IEEE)2024

Multi-Frequency Tracking via Group-Sparse Optimal Transport

Isabel Haasler

In this letter, we introduce an optimal transport framework for inferring power distributions over both spatial location and temporal frequency. Recently, it has been shown that optimal transport is a powerful tool for estimating spatial spectra that chang ...

Ieee-Inst Electrical Electronics Engineers Inc2024

Ensemble learning and ground-truth validation of synaptic connectivity inferred from spike trains

Guillaume Romain Obozinski

Probing the architecture of neuronal circuits and the principles that underlie their functional organization remains an important challenge of modern neurosciences. This holds true, in particular, for the inference of neuronal connectivity from large-scale ...

San Francisco2024

Efficient Neural Clustering and Compression of Strings Through Approximate Euclidean Embeddings of the Levenshtein Distance

Marco Mattavelli, Ünsal Öztürk

Efficient information retrieval is a fundamental requirement in a wide array of applications, ranging from bioinformatics to database systems, and critically hinges on the complexity of an underlying distance metric. This work presents a novel approach to ...

Institute of Electrical and Electronics Engineers Inc.2024

Stable parameterization of continuous and piecewise-linear functions

Michaël Unser, Alexis Marie Frederic Goujon, Joaquim Gonçalves Garcia Barreto Campos

Rectified-linear-unit (ReLU) neural networks, which play a prominent role in deep learning, generate continuous and piecewise-linear (CPWL) functions. While they provide a powerful parametric representation, the mapping between the parameter and function s ...

2023

Fusing physics-based and deep learning models for prognostics

Olga Fink

Physics-based and data-driven models for remaining useful lifetime (RUL) prediction typically suffer from two major challenges that limit their applicability to complex real-world domains: (1) the incompleteness of physics-based models and (2) the limited ...

Elsevier2023

Implicit phonetic information modeling for speech emotion recognition

Tilak Purohit

This study investigates the efficacy of utilizing embedding spaces to model phonetic information in emotion utterances for speech emotion recognition. Our approach involves implicit modeling of phone information by deriving phone-based embeddings from netw ...

Isca-Int Speech Communication Assoc2023

Learning physics-consistent particle interactions

Olga Fink, David Simon Kammer

Interacting particle systems play a key role in science and engineering. Access to the governing particle interaction law is fundamental for a complete understanding of such systems. However, the inherent system complexity keeps the particle interaction hi ...

2022

Machine-learning accelerated identification of exfoliable two-dimensional materials

Kumar Varoon Agrawal, Giovanni Pizzi, Mohammad Tohidivahdat

Two-dimensional (2D) materials have been a central focus of recent research because they host a variety of properties, making them attractive both for fundamental science and for applications. It is thus crucial to be able to identify accurately and effici ...

IOP Publishing Ltd2022

Stochasticity helps to navigate rough landscapes: comparing gradient-descent-based algorithms in the phase retrieval problem

Lenka Zdeborová

In this paper we investigate how gradient-based algorithms such as gradient descent (GD), (multi-pass) stochastic GD, its persistent variant, and the Langevin algorithm navigate non-convex loss-landscapes and which of them is able to reach the best general ...

2021