Cours « Modélisation en neurosciences – et ailleurs » Validation, année 2020-2021 (English instructions below). Une présence régulière au cours est demandée pour être autorisé à valider le cours. Modalités :Choix d'un (ou plusieurs) article(s) en rapport avec le cours. En fonction des intérêts de l'étudiant, ce choix peut résulter d'une proposition de sa part ou de la mienne. Dans tous les cas je dois avoir donné mon accord sur le choix.
Pour le choix d'un article, à titre indicatif, voir les articles qui s(er)ont cités dans les slides ou listés comme supports de cours, ici, et voir les suggestions d'articles faites en 2017, ici (page qu'il me faudrait mettre à jour...), et les articles choisis pour la validation aux printemps 2018 et 2020. A partir du (des) articles(s) lu(s) : présentation (critique) + contribution personnelle (simulation numérique, analyse plus en profondeur d'un aspect mathématique...) NB : 'contribution personnelle', donc projet individuel (pas de travail en groupe). Rapport écrit (sur l'article et le travail personnel) ~ 10 pages - en français ou en anglais ; à rendre au plus tard la veille de l'oral, sous forme électronique (jean-pierre.nadal 'at' phys.ens.fr). Présentation orale : en distanciel, 20mn, + 10 à 15mn de questions sur l'exposé et sur le cours. Autorisé : usage du tableau, de slides, selon besoins. En cas de vidéo-projection, me donner une copie du document (pdf, power-point ou équivalent). Jours/heures de passage : ci-dessous. Instructions in English. Regular attendance is mandatory to get credits from this Course Modalities Choice of one (or possibly several) paper(s) on a topic related to the Course. Contact me to discuss this choice, specifying the particular topics of interest to you, and/or proposing a particular paper. In any case you must obtain my agreement for the paper.
For the choice of an article, as an indication, see the articles that will be listed as course materials, here, the suggestions given in 2017, here (page that I should update...), and the articles chosen for validation in the spring of 2018 and 2020. What you have to do: a critical reading of the paper, + personal micro-project directly related to the paper (numerical simulation, additional mathematical analysis...). Write a report of about 10 pages, and give an oral presentation, on the paper and your personal contribution. The report should be sent to me no later than the day before the oral exam. Oral presentation : video-presentation, 20mns, plus 10-15mn for questions on the paper, your work, as well as on the Course. You can make use of a black/white board, and of slides. In that case, send me an electronic copy of your slides. Connexion link will be sent the day before. Please connect 5 to 10mn in advance so that we can be ontime. Schedule (update: 15 March) ------------------------- Tuesday, March 16 morning, 9:30am-12am 9:30 Alexandre Bois "A compressed sensing perspective of hippocampal function" 10:15 Maria Cherifa "Agglomerative Information Bottleneck" 11:15 Léo Heidelberg "Synaptic Sampling: A Bayesian Approach to Neural Network Plasticity and Rewiring" --------------------------- Thursday, March 18 morning, 9:30am-12:00am 9:30 Charlotte Perlant "The information bottleneck method" 10:15 Auriane Riou "Seeing it all: Convolutional network layers map the function of the human visual system" 11:15 Margaux Tornqvist "Seeing it all: Convolutional network layers map the function of the human visual system" afternoon, 2pm-4:30pm 2:00 Pauline de Lavallade "Network analysis of whole brain fMRI dynamics: Anew framework based on dynamic communicability" 2:45 Tom Dupuis "Convolutional network layers map the function of the human visual system, Thirion et al, 2016" --------------------------- Thursday, March 25 morning, 9:15am-12:30am 9:15 David Boudin "Mutual information, Fisher information, and efficient coding" Wei & Stocker 10:00 Gautier Hamon "Dopamine and temporal difference learning: A fruitful relationship between neuroscience and IA" 11:00 Nicolas Makarof "Seeing it all: Convolutional network layers map the function of the human visual system " afternoon, 2pm 2:00pm Dorian Desblancs "Efficient coding of natural sounds" --------------------------- Thursday, April 8 morning, 9:15am-12:30am 9:15 Etienne Boisseau "Statistics of cone responses to natural images: implications for visual coding", Ruderman et al. 10:00 Clément Bonnet "Principles for biologically based computational models of cortical cognition", O'Reilly, 1999 11:00 Antoine Marion "Efficient Coding of Natural Sounds", Lewicki 11:45 Daniil Lotkov "Unsupervised learning of digit recognition using spike-timing-dependent plasticity", Diehl P. and Cook M. afternoon, 2pm-4:30pm 2:00 Charbel-Raphael Segerie "Brain-Score: Which Artificial Neural Network for Object Recognition is most Brain-Like?" 2:45 Arthur Pellegrino "Engineering recurrent neural networks from task-relevant manifolds and dynamics" Pollock, Jazayeri --------------------------- Week of April 12 Nathan Godey Virgile Dine "Huth et al, A Continuous Semantic Space Describes the Representation of Thousands of Object and Action Categories across the Human Brain