Publications related to Subspace topology

Beyond Spectral Gap: The Role of the Topology in Decentralized Learning

Martin Jaggi, Thijs Vogels, Hadrien Hendrikx

In data-parallel optimization of machine learning models, workers collaborate to improve their estimates of the model: more accurate gradients allow them to use larger learning rates and optimize faster. In the decentralized setting, in which workers commu ...

Microtome Publishing2023

Multi-agent reinforcement learning with graph convolutional neural networks for optimal bidding strategies of generation units in electricity markets

Olga Fink, Mina Montazeri

Finding optimal bidding strategies for generation units in electricity markets would result in higher profit. However, it is a challenging problem due to the system uncertainty which is due to the lack of knowledge of the strategies of other generation uni ...

PERGAMON-ELSEVIER SCIENCE LTD2023

D-Cliques: Compensating for Data Heterogeneity with Topology in Decentralized Federated Learning

Anne-Marie Kermarrec, Erick Lavoie

The convergence speed of machine learning models trained with Federated Learning is significantly affected by non-independent and identically distributed (non-IID) data partitions, even more so in a fully decentralized setting without a central server. In ...

2022

Interaction of lightning-generated electromagnetic fields with overhead and underground cables

Farhad Rachidi-Haeri, Marcos Rubinstein, Carlo Alberto Nucci

In this chapter, we discussed the TL theory and its application to the problem of lightning electromagnetic field coupling to transmission lines. After a short discussion on the underlying assumptions of the TL theory, we described seemingly different but ...

IET2022

A Modular Multilevel Converter with a Clamping Switch for Quasi-Three-Level Operation

Jakub Kucka

The required module capacitances of modular multilevel converters can be significantly reduced in comparison to conventional operation modes when quasi-two-level operation is applied. Unfortunately, the output voltage distortion becomes similar to those of ...

IEEE2021

Entanglement Entropy from Nonequilibrium Work

Jonathan D'Emidio

The Renyi entanglement entropy in quantum many-body systems can be viewed as the difference in free energy between partition functions with different trace topologies. We introduce an external field lambda that controls the partition function topology, all ...

AMER PHYSICAL SOC2020

Convergences of Regularized Algorithms and Stochastic Gradient Methods with Random Projections

Volkan Cevher, Junhong Lin

We study the least-squares regression problem over a Hilbert space, covering nonparametric regression over a reproducing kernel Hilbert space as a special case. We rst investigate regularized algorithms adapted to a projection operator on a closed subspace ...

2020

Experimental studies of the equilibrium and out-of-equilibrium electronic structure of non-symmorphic topological materials

Gianmarco Gatti

Symmetry and topology are fundamental properties of nature. Mathematics provides us with a general framework to understand these concepts. On one side, symmetry describes the invariance properties of an object for specific transformations. On the other sid ...

EPFL2020

Percolation and Internet Science

Emanuele Massaro

Percolation, in its most general interpretation, refers to the “flow” of something (a physical agent, data or information) in a network, possibly accompanied by some nonlinear dynamical processes on the network nodes (sometimes denoted reaction–diffusion s ...

2019

Robust optimization of power grid with distributed generation and improved reliability

Amarasinghage Tharindu Dasun Perera, Zhengchao Wang

Uncertainties of demand, generation and component outage can lead to black-outs in the grid. This study proposes an optimization algorithm (based on Benders Decomposition (BD)) to optimize a Power Transmission System (PTS) that is robust to aforementioned ...

ELSEVIER SCIENCE BV2019

Learning Approach to Delineation of Curvilinear Structures in 2D and 3D Images

Agata Justyna Mosinska

Detection of curvilinear structures has long been of interest due to its wide range of applications. Large amounts of imaging data could be readily used in many fields, but it is practically not possible to analyze them manually. Hence, the need for automa ...

EPFL2019

Long-term evolution of solid oxide fuel and electrolysis cell 3-D microstructure

Giorgio Rinaldi

Solid oxide fuel cells (SOFC) are highly efficient electrochemical conversion devices for the production of electricity that combine fuel flexibility with low emissions of pollutants. The reverse operation in electrolysis mode (SOEC) allows storing electri ...

EPFL2019

Medium Voltage Impedance-Admittance Measurement System Based on the Cascaded H-Bridge Multilevel Converter

Drazen Dujic, Francisco Daniel Freijedo Fernández, Marko Petkovic, Nicolai Hildebrandt

Recent trends in power system design such as an increasing share of renewable energy sources and smart grids are creating different subsystem interactions that require proper investigation, understanding, describing and estimating global system stability t ...

2018

Development of topological tools for the analysis of biological data

Rachel Marcone

There is a growing need for unbiased clustering algorithms, ideally automated to analyze complex data sets. Topological data analysis (TDA) has been used to approach this problem. This recent field of mathematics discerns characteristic features of a space ...

EPFL2018

Neuronal morphologies: the shapes of thoughts

Lida Kanari

The mammalian brain, one of the most fascinating systems in nature, is a complex biological structure that has kept scientists busy for over a century. Many of the brain's mysteries have been unraveled due to the enormous efforts of the scientific communit ...

EPFL2018

Coevolution of Synchronization and Cooperation in Costly Networked Interactions

Alessio Vincenzo Cardillo, Alberto Antonioni

Despite the large number of studies on synchronization, the hypothesis that interactions bear a cost for involved individuals has seldom been considered. The introduction of costly interactions leads, instead, to the formulation of a dichotomous scenario i ...

American Physical Society2017

Effects of Pore and Cage Topology on the Thermodynamics of n-Alkane Adsorption at Bronsted Protons in Zeolites at High Temperature

Berend Smit

Monte Carlo simulations are used to systematically investigate the effects of structural topology on the thermodynamics of n-alkanes adsorbed at Bronsted protons in zeolites having one-dimensional channel systems. In zeolites without cages, the enthalpy an ...

American Chemical Society2017

General Surface Energy for Spinal Cord and Aorta Segmentation

Michaël Unser, Daniel Andreas Schmitter, Virginie Sophie Uhlmann, Harshit Gupta

We present a new surface energy potential for the segmentation of cylindrical objects in 3D medical imaging using parametric spline active contours (a.k.a. spline-snakes). Our energy formulation is based on an optimal steerable surface detector. Thus, we c ...

IEEE2017

Energetic Macroscopic Representation (EMR) and control scheme for the asymmetric 4-stage MSBA

Alfred Rufer, Gina Kristin Steinke

The Multistage Stacked Boost Architecture (MSBA) is a non-isolated step-up converter topology intended for renewable energy generation systems or distribution networks. In previous papers the functionality and intended applications for this topology have b ...

2016

Effects of Zeolite Structural Confinement on Adsorption Thermodynamics and Reaction Kinetics for Monomolecular Cracking and Dehydrogenation of n-Butane

Berend Smit

The effects of zeolite structure on the kinetics of n-butane monomolecular cracking and dehydrogenation are investigated for eight zeolites differing in the topology of channels and cages. Monte Carlo simulations are used to calculate enthalpy and entropy ...

American Chemical Society2016