arXiv subjects

Michaël Fanuel

Publications and source records attributed to Michaël Fanuel.

27 records · Page 2Linked to original sources

arXiv2019

Robust Classification of Graph-Based Data

A graph-based classification method is proposed for semi-supervised learning in the case of Euclidean data and for classification in the case of graph data. Our manifold learning technique is based on a convex optimization problem involving a convex quadratic regularization term and a concave quadratic loss function with a trade-off parameter carefully chosen so that the objective function remains convex. As shown empirically, the advantage of considering a concave loss function is that the learning problem becomes more robust in the presence of noisy labels. Furthermore, the loss function considered here is then more similar to a classification loss while several other methods treat graph-based classification problems as regression problems.

cs.LG↗

arXiv2017

Deformed Laplacians and spectral ranking in directed networks

A deformation of the combinatorial Laplacian is proposed, consisting in a generalization of several existing Laplacians. As particular cases of this construction, the dilation Laplacians are shown to be useful tools for ranking in directed networks of pairwise comparisons. The eigenvector with the smallest eigenvalue of the dilation Laplacians has the same sign on any connected graph, and provides directly a ranking score of its nodes. The ranking method, phrased in terms of a group synchronization problem, is applied to artificial and real data, and its performance is compared with other ranking strategies. A main feature of this approach is the presence of a deformation parameter enabling the emphasis of the top-$k$ objects in the ranking. Furthermore, inspired by these results, a family of random walks interpolating between the undirected random walk and the Pagerank random walk is also proposed.

physics.soc-ph↗

arXiv2016

Magnetic Eigenmaps for the Visualization of Directed Networks

We propose a framework for the visualization of directed networks relying on the eigenfunctions of the magnetic Laplacian, called here Magnetic Eigenmaps. The magnetic Laplacian is a complex deformation of the well-known combinatorial Laplacian. Features such as density of links and directionality patterns are revealed by plotting the phases of the first magnetic eigenvectors. An interpretation of the magnetic eigenvectors is given in connection with the angular synchronization problem. Illustrations of our method are given for both artificial and real networks.

cs.SI↗

arXiv2016

Convex Formulation for Kernel PCA and its Use in Semi-Supervised Learning

In this paper, Kernel PCA is reinterpreted as the solution to a convex optimization problem. Actually, there is a constrained convex problem for each principal component, so that the constraints guarantee that the principal component is indeed a solution, and not a mere saddle point. Although these insights do not imply any algorithmic improvement, they can be used to further understand the method, formulate possible extensions and properly address them. As an example, a new convex optimization problem for semi-supervised classification is proposed, which seems particularly well-suited whenever the number of known labels is small. Our formulation resembles a Least Squares SVM problem with a regularization parameter multiplied by a negative sign, combined with a variational principle for Kernel PCA. Our primal optimization principle for semi-supervised learning is solved in terms of the Lagrange multipliers. Numerical experiments in several classification tasks illustrate the performance of the proposed model in problems with only a few labeled data.

cs.LG↗

arXiv2016

Magnetic eigenmaps for community detection in directed networks

Communities in directed networks have often been characterized as regions with a high density of links, or as sets of nodes with certain patterns of connection. Our approach for community detection combines the optimization of a quality function and a spectral clustering of a deformation of the combinatorial Laplacian, the so-called magnetic Laplacian. The eigenfunctions of the magnetic Laplacian, that we call magnetic eigenmaps, incorporate structural information. Hence, using the magnetic eigenmaps, dense communities including directed cycles can be revealed as well as "role" communities in networks with a running flow, usually discovered thanks to mixture models. Furthermore, in the spirit of the Markov stability method, an approach for studying communities at different energy levels in the network is put forward, based on a quantum mechanical system at finite temperature.

physics.soc-ph↗

arXiv2014

Affine Quantization and the Initial Cosmological Singularity

The Affine Coherent State Quantization procedure is applied to the case of a FRLW universe in the presence of a cosmological constant. The quantum corrections alter the dynamics of the system in the semiclassical regime, providing a potential barrier term which avoids all classical singularities, as already suggested in other models studied in the literature. Furthermore the quantum corrections are responsible for an accelerated cosmic expansion. This work intends to explore some of the implications of the recently proposed "Enhanced Quantization" procedure in a simplified model of cosmology.

hep-th↗

arXiv2014

Non-Perturbative Dynamics, Pair Condensation, Confinement and Dynamical Masses in Massless QED2+1

Quantum electrodynamics in three spacetime dimensions, with one massless fermion species, is studied using a non-perturbative variational approach. Quantization of the theory follows Dirac's Hamiltonian procedure, with a gauge invariant factorization of the physical degrees of freedom. Due to pair condensation in the vacuum state, the symmetry of parity is spontaneously broken. As a consequence, fermionic quasi-particles propagating in the condensate can be identified and are seen to possess a confining dynamical mass, while the propagating physical electromagnetic mode also acquires a non-vanishing dynamical mass. The issues of gauge invariance and confinement of the constituent fermions are carefully discussed.

hep-th↗

arXiv2014

The N = 1 Supersymmetric Wong Equations and the Non-Abelian Landau Problem

A Lagrangian formulation is given extending to N = 1 supersymmetry the motion of a charged point particle with spin in a non-abelian external field. The classical formulation is constructed for any external static non-abelian SU(N) gauge potential. As an illustration, a specific gauge is fixed enabling canonical quantization and the study of the supersymmetric non-abelian Landau problem. The spectrum of the quantum Hamiltonian operator follows in accordance with the supersymmetric structure.

hep-th↗

arXiv2011

Dressed Fermions, Modular Transformations and Bosonization in the Compactified Schwinger Model

The celebrated exactly solvable "Schwinger" model, namely massless two-dimensional QED, is revisited. The solution presented here emphasizes the non- perturbative relevance of the topological sector through large gauge transformations whose role is made manifest by compactifying space into a circle. Eventually the well-known non-perturbative features and solution of the model are recovered in the massless case. However the fermion mass term is shown to play a subtle role in order to achieve a physical quantization that accounts for gauge invariance under both small and large gauge symmetries. Quantization of the system follows Dirac's approach in an explicitly gauge invariant way that avoids any gauge fixing procedure.

hep-th↗