Publications associées à Loi triangulaire

On the symmetries in the dynamics of wide two-layer neural networks

We consider the idealized setting of gradient flow on the population risk for infinitely wide two-layer ReLU neural networks (without bias), and study the effect of symmetries on the learned parameters and predictors. We first describe a general class of s ...

AMER INST MATHEMATICAL SCIENCES-AIMS2023

A Theory of Finite-Width Neural Networks: Generalization, Scaling Laws, and the Loss Landscape

Berfin Simsek

Deep learning has achieved remarkable success in various challenging tasks such as generating images from natural language or engaging in lengthy conversations with humans. The success in practice stems from the ability to successfully train massive neural ...

EPFL2023

Dalton: Learned Partitioning for Distributed Data Streams

Anastasia Ailamaki, Eleni Zapridou, Ioannis Mytilinis

To sustain the input rate of high-throughput streams, modern stream processing systems rely on parallel execution. However, skewed data yield imbalanced load assignments and create stragglers that hinder scalability. Deciding on a static partitioning for a ...

2022

From Trees to Barcodes and Back Again: A Combinatorial, Probabilistic and Geometric Study of a Topological Inverse Problem

Adélie Eliane Garin

In this thesis, we investigate the inverse problem of trees and barcodes from a combinatorial, geometric, probabilistic and statistical point of view. Computing the persistent homology of a merge tree yields a barcode B. Reconstructing a tree from B invol ...

EPFL2022

Iron incorporation in synthetic precipitated calcium silicate hydrates

David Lyndon Emsley, Paul Bowen, Brennan James Walder, Jirawan Siramanont

Effect of Fe3+ on the precipitation of synthetic calcium silicate hydrates (C-S-H) under controlled conditions has been evaluated. Using extremely basic either initial calcium nitrate or sodium silicate solutions (pH > 11), incipient formation of ferrihydr ...

Pergamon-Elsevier Science Ltd2021

The probability of intransitivity in dice and close elections

Jan Hazla

We study the phenomenon of intransitivity in models of dice and voting. First, we follow a recent thread of research for n-sided dice with pairwise ordering induced by the probability, relative to 1/2, that a throw from one die is higher than the other. We ...

2020

Subjective and physiological responses to façade and sunlight pattern geometry in virtual reality

Marilyne Andersen, Jan Wienold, Giorgia Chinazzo, Kynthia Chamilothori, João Pedro de Matos Rodrigues

This study investigates the joint impact of façade geometry and associated sunlight patterns on occupant subjective perception and physiological responses through a novel experimental method coupling physically-based simulations shown in virtual reality wi ...

2019

Particle number control for direct simulation Monte-Carlo methodology using kernel estimates

Mohammadhossein Gorji

The efficiency of stochastic particle schemes for large scale simulations relies on the ability to preserve a uniform distribution of particles in the whole physical domain. While simple particle split and merge algorithms have been considered previously, ...

2019

Statistical Aspects of Wasserstein Distances

Victor Panaretos, Yoav Zemel

Wasserstein distances are metrics on probability distributions inspired by the problem of optimal mass transportation. Roughly speaking, they measure the minimal effort required to reconfigure the probability mass of one distribution in order to recover th ...

ANNUAL REVIEWS2019

On the Limits of Precision Glass Molding for Diffractive Optical Elements

Karin Prater

Diffractive optical elements (DOEs) consist of surface reliefs with dimensions in the micrometer range and nanometer precision. Two technologies dominate: Elements replicated in plastic and directly microfabricated elements in fused silica. Plastic DOEs ar ...

EPFL2017

Replenishment of sediment downstream of dams : erosion and transport processes

Elena Battisacco

Dams on rivers alter the sediment continuum, trapping water and sediment in the upstream reservoirs. River reaches downstream are affected by several negative effects, such as bed incision, reduction of the river morphological variability, development of a ...

EPFL2016

Re-proving Channel Polarization Theorems : An Extremality and Robustness Analysis

Mine Alsan

The general subject considered in this thesis is a recently discovered coding technique, polar coding, which is used to construct a class of error correction codes with unique properties. In his ground-breaking work, Arikan proved that this class of codes, ...

EPFL2015

Extremal Channels of Gallager's E-0 Under the Basic Polarization Transformations

Mine Alsan

We study the extremality of the binary erasure channel and the binary symmetric channel for Gallager's reliability function E-0 of binary input discrete memoryless channels evaluated under the uniform input distribution from the aspect of channel polarizat ...

Institute of Electrical and Electronics Engineers2014

Polar Codes for the Two-User Multiple-Access Channel

Emre Telatar, Eren Sasoglu

Arikan's polar coding method is extended to two-user multiple-access channels. It is shown that if the two users of the channel use Arikan's construction, the resulting channels will polarize to one of five possible extremals, on each of which uncoded tran ...

Institute of Electrical and Electronics Engineers2013

Extended Uniform Distribution Accounting for Uncertainty of Uncertainty

Ian Smith, James-Alexandre Goulet

Increasingly, uncertainties are explicitly considered for important engineering tasks. Often, little case-specific information is available for characterizing these uncertainties. Uniform distributions are an easy way to describe errors in absence of more ...

ASCE2011

Almost-Uniform Sampling of Points on High-Dimensional Algebraic Varieties

Mohammad Amin Shokrollahi, Mahdi Cheraghchi Bashi Astaneh

We consider the problem of uniform sampling of points on an algebraic variety. Specifically, we develop a randomized algorithm that, given a small set of multivariate polynomials over a sufficiently large finite field, produces a common zero of the polynom ...

2009

Regional climate model data used within the SWURVE project. 2: Addressing uncertainty in regional climate model data for five European case study areas

Abdelkader Mezghani

To aid assessments of climate change impacts on water related activities in the case study regions (CSRs) of the EC funded project SWURVE, estimates of uncertainty in climate model data need to be developed. In this paper, two methods to estimate uncertain ...

2007