TNG Team Member

Emmanuel Dauce

Emmanuel Dauce, MCU
TNG


Email:

Phone number:
+33 4 91 29 98 19



Emmanuel Daucé is associate professor at the Ecole Centrale de Marseille, doing his research in Computational Neuroscience at the "Institut de Neurosciences des Systèmes" (France), a joint research unit (Inserm/ CNRS / Aix-Marseille Université). His research lies at the crosssroad of machine learning, artificial intelligence and neuroscience, seeking to develop innovative computational models and methods though remaining consistent with the principles of biological systems. 

He graduated from the Ecole Nationale Supérieure d'Electronique, d'Electrotechnique, d'Informatique et d'Hydraulique de Toulouse (1995), and obtained a Ph.D in Knowledge Representation and Formal reasoning from the Ecole Nationale Supérieure de l'Aeronautique et de l'Espace (2000), under the supervision of Bernard Doyon (Inserm) and Manuel Samuelides (ISAE), on learning and plasticity in artificial neural networks with random recurrent connectivity graphs ( Daucé et al, 1998 ). He contributed to extend the model to multiple populations (Daucé et al. ,2001), and spatio-temporal sequence learning (Daucé et al., 2002).
He joined the Institut des Sciences du Mouvement in Marseille in 2001, where he contributed to develop neurally plausible reinforcement schemes in closed-loop control systems (Daucé, 2004), address spike-timing dependent plasticity in balanced networks of spiking neurons (Henry et al, 2006) and develop models of dynamic retention in discrete neural-fields (Daucé, 2004).
He more recently joined Viktor Jirsa's group at the Institut de Neurosciences des Systèmes, at the Faculté de Médecine de La Timone (Marseille), where he contributed to develop on-line learning methods for non-stationary data streams - adapted to the case of Brain Computer Interfaces (Daucé and Thomas, 2014), and participated in modelling brain non-stationarities with simple neural-mass dynamics on large-scale connectivity graphs (Golos et al., 2015).

--> personal website

HDR : apprentissage et contrôle dans les architectures neuronales

The brain, beyond its primary sensori-motor and regulation functions, is an outstanding adaptive system, capable of developping novel responses in novel situations. The principles of machine learning, a fast-developping domain, are at stake for a better understanding of the learning processes in the brain. Computational models of learning have provided several success stories, from which the "layered neural networks" are the most famous ones. This HDR dissertation presents different kinds neural networks models, displaying a more strict obedience to the biological constraints, in particular regarding the recurrent aspect of the neuronal interaction graph, the discreteness of the signals emitted by the neurons and the local aspect of the plasticity rules that govern the synaptic changes. We show in particular how recurrent neural networks organize their sensory input in different regions, how the the synaptic plasticity drives the network toward a more "simple" collective activity, allowing a better separation and prediction of the sensory stimuli, and how motor learning can rely on matching motor primitives with sensory data to organize the physical environment. Several projects are proposed, aiming at expanding some of those ideas into large-scale brain activity models, or also for the design of brain-computer interfaces.

 

Scientific activity since 1995

 

    • Daucé, E. (1995) Dynamique chaotique et apprentissage au sein de réseaux neuromimétiques, rapport de DEA Représentation de la Connaissance et Formalisation du Raisonnement, 26 juin 1995, Université Paul Sabatier, Toulouse, France.
    • Cessac, B., Quoy, M., Daucé, E. Doyon, B. and Samuelides, M. (1996) Learning in a chaotic neural network, proc. of workshop on Neural Network Dynamics and Pattern Recognition (DYNN'96), 12-13 mars, Toulouse, France.

    • Daucé, E. and Doyon, B. (1998) Un processus d'apprentissage hebbien dans des réseaux dynamiques à temps discret, actes des Neuvièmes journées Neurosciences et Sciences de l'Ingénieur (NSI), may 11-14, Munster, France.
    • Daucé E. and Doyon, B. (1998) Novelty Learning in a Discrete Time Chaotic Network, proc. of the 8th International Conference on Artificial Neural Networks (ICANN'98), L. Niklasson et al. eds, Vol. 2: 1051-1056, Springer, September 2-4, Skövde, Sweden.

    • Daucé, E., Moynot, O., Pinaud, O., Samuelides, M. and Doyon, B. (1999) Mean Field Equations reveal synchronisation in a 2-population neural network model, proc. of the 7th European Symposium On Artificial Neural Networks (ESANN'99), Verleysen, M. ed.:7-12, April 21-23, Bruges, Belgium.
    • Moynot, O., Daucé, E. and Pinaud, O. (1999) Equations de champ moyen pour les réseaux de neurones à deux populations, actes du IIIème Colloque Jeunes Chercheurs en Sciences Cognitives, B. Vivicorsi et al. eds, 26-28 Avril, Soulac, France.
    • Daucé, E. and Doyon, B. (1999) Apprentissage dynamique dans les réseaux de neurones, actes du IIIème Colloque Jeunes Chercheurs en Sciences Cognitives, B. Vivicorsi et al. eds, 26-28 Avril, Soulac, France.
    • Moynot, O., Daucé, E. and Pinaud, O. (1999) Equations de champ moyen pour les réseaux de neurones à deux populations, In Cognito 15:41-46.
    • Daucé, E. (2000) Adaptation dynamique et apprentissage dans des réseaux de neurones récurrents aléatoires, thèse de doctorat, 13 janvier 2000, ENSAE, Toulouse, France.
    • Daucé, E. and Doyon, B. (2000) Un modèle neuronal de la reconnaissance de scènes dynamiques, journée thématique PRESCOT apprentissage dans les systèmes naturels et artificiels, 5 mai 2000, LAAS, Toulouse, France.
    • Quoy M. and Daucé, E. (2000) Visual and motor learning using a chaotic recurrent neural network: application to the control of a mobile robot, proc. of the  Second International ICSC Symposium on Neural Computation (NC2000), Bohte, H. and Rojas, R. eds: 577-582, May 23-26, Berlin, Germany.
    • Daucé, E. (2000) Apprentissage et reconnaissance de signaux spatio-temporels sur des réseaux neuronaux à dynamiques complexes, VIIIe Journée d'Etude ACCION : théorie du chaos et sciences de la cognition : Chimère ou Réalité, 29-30 Juin, Aix-en Provence, France.
    • Daucé E. and Quoy, M. (2000) Random recurrent neural networks for autonomous systems design, additional, proc. of the sixth international conference on Simulation of Adaptive Behavior : From Animals to Animats (SAB 2000), Meyer, J.-A. et al. eds:31-40, Sept. 11-15, Paris, France.
    • Daucé, E. (2000) Dynamical memories on large recurrent neural networks, proc. of international workshop on Dynamical Neural Networks and applications (DYNN'2000), Nov. 20-25, Bielefeld, Germany.


    • Daucé, E., Quoy, M. and Doyon, B. (2002) Resonant spatio-temporal learning in large random recurrent networks, Biol.Cybern. 87(3):185-198.
    • Guillot, A. and Daucé, E. éds (2002) Approche dynamique de la cognition artificielle, Lavoisier, Paris, France.
    • Daucé, E. (2002) Systèmes dynamiques pour les sciences cognitives, in Approche dynamique de la cognition artificielle, Guillot, A. and Daucé, E. eds :33-44, Lavoisier, Paris, France.
    • Beslon, G. and Daucé, E. (2002) Modularité et apprentissage dans les réseaux de neurones récurrents, in Approche dynamique de la cognition artificielle, Guillot, A. and Daucé, E. eds :61-80, Lavoisier, Paris, France.
    • Daucé, E. and Quoy, M. (2002) Mémoire dynamique et planification, in Approche dynamique de la cognition artificielle, Guillot, A. and Daucé, E. eds :97-113, Lavoisier, Paris, France.
    • Daucé, E and Guillot, A. (2002), Conclusion, in Approche dynamique de la cognition artificielle, Guillot, A. and Daucé, E. eds :303-306, Lavoisier, Paris, France.
    • Daucé, E (2003) Dynamics of Neural Networks II - simulation and learning , Journées post-génomiques de la Douai (JPGD), may 14-16, Lyon, France.
    • Daucé, E. (2003) Dynamic Retention in Recurrent Networks of Spiking Neurons, proc. of the workshop on the FUture of Neural Networks (FUNN 2003), Thirtieth International Colloquium on Automata, Languages and Programming, June 30 - July 4, Eindhoven, The Netherlands.
    • Daucé, E. (2004) Local and global causality in biomimetic systems, journées d'étude Emergentisme dynamique et fondement des sciences de la cognition, séminaires d'épistémologie des sciences cognitives, ENS-LSH,  22-24 avril 2004, Lyon, France.
    • Daucé, E. (2004) Short term memory in recurrent networks of spiking neurons, Natural Computing 3:135-157
    • Daucé, E. (2004) Différentes méthodes d'apprentissage dans les réseaux récurrents, atelier "Neurosciences intégratives et computationnelles" , ACI "Temps et cerveau", 16-18 juin, CIRM, Luminy, Marseille, France.
    • Daucé, E. (2004) Hebbian reinforcement learning in a modular dynamic network, proc. of the eighth international conference on Simulation of Adaptive Behavior : From Animals to Animats (SAB*04), Schaal, S. et al. eds:305-314, July 13-17, Los Angeles, CA, USA.
    • Daucé, E. (2004) The dynamics of appropriation, proc. of the 2nd workshop on Anticipatory Behavior in Adaptive Learning Systems 2004 (ABiALS 2004), July 17th 2004, Los Angeles, CA, USA.


    • Daucé, E. (2005) Loi de Hebb dans des réseaux aléatoires multi-populations, atelier "Neurosciences intégratives et computationnelles" , ACI "Temps et cerveau", 6-8 juin 2005, Ile Ste Marguerite, Cannes, France.
    • Daucé, E., Soula, H. and Beslon, G. (2005) Learning Methods for Dynamic Neural Networks, proc. of the 2005 International Symposium on Nonlinear Theory and its Applications (NOLTA'05):598-601, Oct. 18-21, Bruges, Belgium.
    • Daucé, E. (2005) Dynamics and learning in large random neural networks, Riken Brain Science Institute, october 5, 2005, Saitama, Japan.
    • Daucé, E. (2006) Apprentissage sensori-moteur sur des réseaux de neurones récurrents aléatoires, Journée "Neurones et réseaux de neurones biologiques : comment les modéliser?", journée du GDR MSPC, 20 janvier 2006, Institut Henri Poincaré, Paris, France.
    • Henry, F., Daucé, E. and Soula, H. (2006) Temporal Pattern Identification using Spike-Timing Dependent Plasticity (2006), proc. of  the 15th annual Computational Neurosciences meeting (CNS*2006), July 16-18, Edinburgh, U.K.

    • Daucé, E. and Henry, F. (2006) Hebbian Learning in Large Recurrent Neural Networks, proc. of 1ère conférence française des neurosciences computationnelles (Neurocomp'06), Alexandre F. et al. eds : 202-205, Oct. 23-24 , Pont-à-Mousson, France.
    • Henry, F., Daucé, E. and Soula, H. (2006) Temporal Pattern Identification using Spike Timing Dependent Plasticity, Neurocomputing 70(10-12): 2009-2016.
    • Daucé, E. (2007)  Hebbian learning in large recurrent neural networks, séminaire des laboratoires du Pôle3C, 30 mars 2007, Université de Provence, Marseille, France.
    • Daucé, E. and Henry, F. (2007) STDP-induced periodic encoding of static patterns in balanced recurrent neural networks, unpublished.
    • Daucé, E. (2007) Learning and Control with large Dynamic Neural Networks, European Physical Journal - Special Topics 142:123-161.
    • Samuelides, M., Cessac, B., Beslon, G., Daucé, E., Perrinet, L., Quoy, M. and Thorpe, S. (2007) Dynamique des réseaux de neurones artificiels biologiquement plausibles en robotique autonome, Colloque de l'ACI "Neurosciences intégratives et computationnelles", 11-12 juin 2007, Collège de France, Paris, France.

    • Henry, F. and Daucé, E. (2008)  Emergence of stimulus-specific synchronous response  through STDP in recurrent neural networks, proc. of the 16th European Symposium on Artificial Neural Networks - Advances in  Computational Intelligence and Learning (ESANN 2008), Verleysen, M. ed: 379-384,   April 23-25, Bruges, Belgium.
    • Henry, F. and Daucé, E. (2008) Spike-Timing Dependent Plasticity and regime transitions in random recurrent neural networks, proc. of the second french conference on computational neurosciences (Neurocomp'08), Daucé, E. and Perrinet, L. eds: 291-295, October 8-11, Marseille, France.
    • Daucé, E. and Perrinet, L.  éds (2008) Proceedings of the second French conference on Computational Neuroscience (Neurocomp08), October 8-11, Marseille, France, ISBN 978-2-9532965-0-1.
    • Daucé, E. (2008) Le projet MAPS : intelligence topographique et adaptation comportementale, Journée scientifique SCHEME (Sciences Cognitives en ingénierie des facteurs Humains et ErgonoMiE), 14 octobre 2008, Toulouse, France.