Publication

Stochastic Gradient Descent for Spectral Embedding with Implicit Orthogonality Constraint

Publications associées (32)

Regularization Techniques for Low-Resource Machine Translation

Neural machine translation (MT) and text generation have recently reached very high levels of quality. However, both areas share a problem: in order to reach these levels, they require massive amounts of data. When this is not present, they lack generaliza ...

EPFL2023

Fairness and Explainability in Clustering Problems

Xinrui Jia

In this thesis we present and analyze approximation algorithms for three different clustering problems. The formulations of these problems are motivated by fairness and explainability considerations, two issues that have recently received attention in the ...

EPFL2023

Unsupervised Graph Representation Learning with Cluster-aware Self-training and Refining

Qiang Liu, Feng Yu, Yichen Xu

Unsupervised graph representation learning aims to learn low-dimensional node embeddings without supervision while preserving graph topological structures and node attributive features. Previous Graph Neural Networks (GNN) require a large number of labeled ...

Assoc Computing Machinery2023

A mixed precision LOBPCG algorithm

Daniel Kressner, Meiyue Shao, Yuxin Ma

The locally optimal block preconditioned conjugate gradient (LOBPCG) algorithm is a popular approach for computing a few smallest eigenvalues and the corresponding eigenvectors of a large Hermitian positive definite matrix A. In this work, we propose a mix ...

SPRINGER2023

Fast Adversarial Training With Adaptive Step Size

Mathieu Salzmann, Sabine Süsstrunk, Jue Wang, Chen Liu, Yong Zhang, Zhuoyi Huang

While adversarial training and its variants have shown to be the most effective algorithms to defend against adversarial attacks, their extremely slow training process makes it hard to scale to large datasets like ImageNet. The key idea of recent works to ...

Ieee-Inst Electrical Electronics Engineers Inc2023

Hamiltonian Deep Neural Networks Guaranteeing Non-Vanishing Gradients by Design

Giancarlo Ferrari Trecate, Clara Lucía Galimberti, Liang Xu, Luca Furieri

Deep Neural Networks (DNNs) training can be difficult due to vanishing and exploding gradients during weight optimization through backpropagation. To address this problem, we propose a general class of Hamiltonian DNNs (H-DNNs) that stem from the discretiz ...

2023

Byzantine Fault-Tolerance in Federated Local SGD Under 2f-Redundancy

Nirupam Gupta

In this article, we study the problem of Byzantine fault-tolerance in a federated optimization setting, where there is a group of agents communicating with a centralized coordinator. We allow up to

f

Byzantine-faulty agents, which may not follow a prescr ...

Ieee-Inst Electrical Electronics Engineers Inc2023

espm: A Python library for the simulation of STEM-EDXS datasets

Cécile Hébert, Nathanaël Perraudin, Duncan Thomas Lindsay Alexander, Hui Chen, Guillaume Romain Obozinski

We present two open-source Python packages: "electron spectro-microscopy"(espm) and "electron microscopy tables"(emtables). The espm software enables the simulation of scanning transmission electron microscopy energy-dispersive X-ray spectroscopy datacubes ...

ELSEVIER2023

Probabilistic methods for neural combinatorial optimization

Nikolaos Karalias

The monumental progress in the development of machine learning models has led to a plethora of applications with transformative effects in engineering and science. This has also turned the attention of the research community towards the pursuit of construc ...

EPFL2023

A Multiconfigurational Wave Function Implementation of the Frenkel Exciton Model for Molecular Aggregates

Francesco Aquilante

We present an implementation of the Frenkel exciton model into the OpenMolcas program package enabling calculations of collective electronic excited states of molecular aggregates based on a multiconfigurational wave function description of the individual ...

AMER CHEMICAL SOC2023

Training Efficient Controllers via Analytic Policy Gradient

Dario Floreano, Davide Scaramuzza, Valentin Wüest

Control design for robotic systems is complex and often requires solving an optimization to follow a trajectory accurately. Online optimization approaches like Model Predictive Control (MPC) have been shown to achieve great tracking performance, but requir ...

New York2023

Stable Nonconvex-Nonconcave Training via Linear Interpolation

Volkan Cevher, Thomas Michaelsen Pethick, Wanyun Xie

This paper presents a theoretical analysis of linear interpolation as a principled method for stabilizing (large-scale) neural network training. We argue that instabilities in the optimization process are often caused by the nonmonotonicity of the loss lan ...

2023

Publications associées (32)

f

Byzantine-faulty agents, which may not follow a prescr ...

Ieee-Inst Electrical Electronics Engineers Inc2023