Computational models in neurosciences between mechanistic and phenomenological characterizations - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2022

Computational models in neurosciences between mechanistic and phenomenological characterizations

Résumé

Objective: Computational neuroscience combines mathematics, computer science models, and neurosciences for theorizing, investigating, and simulating neural systems involved in the development, structure, physiology, and cognitive abilities of the brain. Computational models constitute a major stake in translational neuroscience: the analytical understanding of these models seems fundamental to consider a translation towards clinical applications. Method: We propose a minimal typology of computational models, which allows distinguishing between more realistic models (e.g., mechanistic models) and pragmatic models (e.g., phenomenological models). Result: Understanding the translational aspects of computational models goes far beyond the intrinsic characteristics of models. First, we assume that a computational model is rarely uniquely mechanistic or phenomenological. Idealization seems necessary because of i) the researcher's perspectives on the phenomena and the purposes of the study (i.e., by the relativity of the model); ii) The complexity of reality across different levels and therefore the nature and number of dimensions required to consider a phenomenon. Especially, the use of models goes far beyond their function, and requires considering external characteristics rooted in path dependence, interdisciplinarity, and pluralism in neurosciences. Conclusion: The unreasonable use of computational models, which are highly complex and subject to a shift in their initial function, could be limited by bringing to light such factors.
Fichier principal
Vignette du fichier
Model_CG_CB_TB_MC_DD_2022.pdf (386.02 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03508973 , version 1 (03-01-2022)

Identifiants

  • HAL Id : hal-03508973 , version 1

Citer

Gauld Christophe, Cédric Brun, Thomas Boraud, Mallory Carlu, Damien Depannemaecker. Computational models in neurosciences between mechanistic and phenomenological characterizations. 2022. ⟨hal-03508973⟩
65 Consultations
409 Téléchargements

Partager

Gmail Facebook X LinkedIn More