HAL Paris Dauphine-PSL
Not a member yet
21932 research outputs found
Sort by
Face à l’IA, les médias ne pourront pas faire l’économie d’une profonde remise en question de leur mode de fonctionnement
No full textL’intelligence artificielle remet profondément en cause la raison d’être des médias et leur modèle d’affaires. Leur survie pourrait passer par une rapide réflexion sur ce qui fait leur cœur de métier, à l’image de la politique suivie par le New York Times. Au risque de sombrer dans des offres très low cost ou de devenir des fournisseurs des géants de l’IA
Un capital bureaucratique au profit de la finance impériale: La Compagnie générale des colonies, 1920-1959
No full textInternational audienceThis article studies how the appropriation of bureaucratic capital – that is the enrolment of former high and mid-level civil servants, as board members, directors or operational agents – allowed the activation of imperial profits, within one of France’s main financial holdings. The capture of administrative experience, knowhow and interpersonal networks proved key in securing vital resources for colonial affairs: land concessions, public contracts, and forced labor. In studying the business of “pantouflage”, this paper identifies its two main functions – patronage and operational - which respectively ensured the legitimacy of the firm, and allowed the upkeep of good relations with various layers of the colonial administration: from governors to cadastral authorities, from labor and health inspections to public works. In doing so, we contribute to the political economy of colonialism through a better understanding of public-private circulations among imperial economic personnel.Cet article étudie l’appropriation d’un capital bureautique – défini par l’accumulation d’un savoir d’Etat à travers le recrutement d’anciens hauts et moyens fonctionnaires, d’ingénieurs ou de diplômés d’Etat – par une grande firme coloniale, comme condition d’activation de profits impériaux. La mobilisation d’expériences, de savoir-faire administratifs et de réseaux d’interconnaissance a représenté un facteur clé en vue d’accumuler les ressources nécessaires aux affaires coloniales, que sont les concessions, les contrats publics et la main-d’œuvre (contrainte). En s’intéressant à la dimension stratégique du « pantouflage », nous distinguons entre deux fonctions – de parrainage et opérationnelle – qui, respectivement, légitiment le capital financier et en assurent la fructification, par un faisceau de relations avec différentes strates de l’administration : des gouverneurs généraux aux autorités locales, du cadastre à l’inspection du travail, des travaux publics aux services sanitaires. En étudiant la manière dont une firme financière puise dans différents corps administratifs pour conduire ses affaires, ce travail apporte un éclairage neuf sur l’économie politique du colonialisme, saisi à partir des circulations public-privé propres aux élites économiques impériales
Exit Incentives for Carbon Emissive Firms
No full textInternational audienceWe develop a continuous-time model of incentives for carbon emissive firms to exit the market based on a compensation payment identical to all firms. In our model, firms enjoy profits from production modeled as a simple geometric Brownian motion and do not bear any environmental damage from production. A regulator maximizes the expected discounted value of firms profits from production minus environmental damages caused by production and proposes a compensation payment whose dynamics is known to the firms. We provide in both situations closed-form expressions for the compensation payment process and the exit thresholds of each firms. We apply our model to the crude oil market. We show that market concentration increases the total expected discounted payments to firms and reduces the expected closing time of polluting assets. A certain degree of market concentration can enable the regulator to halt production in a shorter time but at a higher cost. We extend this framework to the case of two countries each regulating its own market. We analyze scenarios involving two large producers and a combination of large and small producers. Our findings indicate that the proposed model facilitates an earlier exit of brown energy from the market compared to the scenario where each country independently regulates its market. This has significant implications for regulatory strategies aimed at accelerating the transition to cleaner energy sources
Convergence of Multi-Level Markov Chain Monte Carlo Adaptive Stochastic Gradient Algorithms
No full textStochastic optimization in learning and inference often relies on Markov chain Monte Carlo (MCMC) to approximate gradients when exact computation is intractable. However, finite-time MCMC estimators are biased, and reducing this bias typically comes at a higher computational cost. We propose a multilevel Monte Carlo gradient estimator whose bias decays as while its expected computational cost grows only as , where is the maximal truncation level at iteration n. Building on this approach, we introduce a multilevel MCMC framework for adaptive stochastic gradient methods, leading to new multilevel variants of Adagrad and AMSGrad algorithms. Under conditions controlling the estimator bias and its second and third moments, we establish a convergence rate of order up to logarithmic factors. Finally, we illustrate these results on Importance-Weighted Autoencoders trained with the proposed multilevel adaptive methods
On the Cutoff Phenomenon for Dyson-Jacobi Processes
No full textWe study the convergence to equilibrium of the Dyson-Jacobi process, a system of n interacting particles on the segment [0, 1] arising from Random Matrix Theory. We establish the occurence of a cutoff phenomenon for the intrinsic Wasserstein distance and provide an explicit formula for the associated mixing time.Our approach relies on the interplay between the Riemannian geometry of the process and a flattened Euclidean representation obtained via a diffeomorphic deformation. This transformation allows us to transfer curvature-dimension inequalities from the Euclidean setting to the original space, thereby yielding sharp quantitative estimates.</div
Monotonicity of the period and positive periodic solutions of a quasilinear equation
No full textInternational audienceWe investigate the monotonicity of the minimal period of periodic solutions of quasilinear differential equations involving the p-Laplace operator. First, the monotonicity of the period is obtained as a function of a Hamiltonian energy in two cases. We extend to p≥2 classical results due to Chow-Wang and Chicone for p=2. Then we consider a differential equation associated with a fundamental interpolation inequality in Sobolev spaces. In that case, we generalize monotonicity results by Miyamoto-Yagasaki and Benguria-Depassier-Loss to p≥2
Grouping strategies on two-phase methods for bi-objective combinatorial optimization
No full textInternational audienc
Benign landscape for Burer-Monteiro factorizations of MaxCut-type semidefinite programs
No full textInternational audienceWe consider MaxCut-type semidefinite programs (SDP) which admit a low rank solution. To numerically leverage the low rank hypothesis, a standard algorithmic approach is the Burer-Monteiro factorization, which allows to significantly reduce the dimensionality of the problem at the cost of its convexity. We give a sharp condition on the conditioning of the Laplacian matrix associated with the SDP under which any second-order critical point of the non-convex problem is a global minimizer. By applying our theorem, we improve on recent results about the correctness of the Burer-Monteiro approach on Z 2 -synchronization problems
