FID4SA-Repository
Not a member yet
21181 research outputs found
Sort by
Bayesian social aggregation with accumulating evidence
No full textInternational audienceHow should we aggregate the ex ante preferences of Bayesian agents with heterogeneous beliefs? Suppose the state of the world is described by a random process that unfolds over time. Different agents have different beliefs about the probabilistic laws governing this process. As new information is revealed over time by the process, agents update their beliefs and preferences via Bayes rule. Consider a Pareto principle that applies only to preferences which remain stable in the long run under these updates. I show that this “eventual Pareto” principle implies that the social planner must be a utilitarian. But it does not impose any relationship between the beliefs of the individuals and those of the planner, except for a weak compatibility conditio
Instrumental Genesis in the Theory of MWS: Insight from Didactic Research on Digital Artifacts
No full textInternational audienceThe central focus of this chapter is the relationship between the notions introduced by Vérillon and Rabardel (1995), 'artefact' and 'instrumental genesis', and the homonym notions in the theory of MWS. Following a brief overview of the Piagetian and Activity Theory perspectives which inspired Vérillon and Rabardel, we explore how didactic research on digital artefacts developed two viewpoints on the instrumental genesis: psychological and institutional. The idea of instrumental genesis in the theory of MWS is consistent with the concept of instrumented mediation by Vérillon and Rabardel, but is a more focused notion, since MWS include two additional geneses, semiotic and discursive. The theory of MWS moreover does not theorize about the mediation and as such, in a particular situation where a digital artefact is used, both viewpoints will be helpful. Regarding the poles of the instrumental genesis, we propose characterizing a digital artefact by the underlying data representations and algorithmic treatments, and constructions as data-result models of mathematical configurations, taking advantage of the representations and algorithms in the artefact
Ultrasonic microrheology for ex vivo skin explants monitoring: A proof of concept
No full textInternational audienc
Guerre : « Il faut penser avant de s’indigner » Le laboratoire du bonheur
No full textchronique radio France Ble
AAEGAN Optimization by Purposeful NoiseInjection for the Generation of Bright-Field BrainOrganoid Images
No full textInternational audienceBrain organoids are three-dimensional tissues generatedin vitro from pluripotent stem cells and replicating the earlydevelopment of Human brain. To implement, test and comparemethods to follow their growth on microscopic images, a largedataset not always available is required with a trusted groundtruth when developing automated Machine Learning solutions.Recently, optimized Generative Adversarial Networks prove togenerate only a similar object content but not a backgroundspecific to the real acquisition modality.In this work, a small database of brain organoid bright fieldimages, characterized by a shot noise background, is extendedusing the already validated AAEGAN architecture, and specificnoise or a mixture noise injected in the generator. We hypothesizethis noise injection could help to generate an homogeneous andsimilar bright-field background. To validate or invalidate ourgenerated images we use metric calculation, and a dimensionalreduction on features on original and generated images.Our result suggest that noise injection can modulate thegenerated image backgrounds in order to produce a more similarcontent as produced in the microscopic reality. A validation ofthese images by biological experts could augment the originaldataset and allow their analysis by Deep-based solutions
From the Networking of Theories to the Discussion of the Educational Implications of Research
No full textInternational audienc
Filiations et renouvellement dans les pratiques d’écriture créative en formations
No full textInternational audienc
The Debate on Super-quantum Correlations: Statistical Misinterpretation and Logical Fallacy
No full textInternational audienceThis article shows that the mainstream discussion about super-quantum correlations is skewed by an incorrect interpretation of the “no-signaling” condition (NS). Referring to counter-examples, it shows that the usual probabilistic interpretation of (NS), whose link with relativistic causality is doubtful, is too weak to assert the absence of any exchange of information between the parties. A relevant informal interpretation of (NS) is nothing but a particular specification of Pawlowski’s Information Causality principle, which rules out the possibility of correlations stronger than the strongest quantum correlations