En poursuivant votre navigation sur ce site, vous acceptez le dépôt de cookies dans votre navigateur. (En savoir plus)

PhD student in computer science and mathematics (H/F)

This offer is available in the following languages:
Français - Anglais

Date Limite Candidature : jeudi 2 juin 2022

Assurez-vous que votre profil candidat soit correctement renseigné avant de postuler. Les informations de votre profil complètent celles associées à chaque candidature. Afin d’augmenter votre visibilité sur notre Portail Emploi et ainsi permettre aux recruteurs de consulter votre profil candidat, vous avez la possibilité de déposer votre CV dans notre CVThèque en un clic !

General information

Reference : UMR7271-VIVROS-031
Workplace : SOPHIA ANTIPOLIS
Date of publication : Thursday, May 12, 2022
Scientific Responsible name : Alexandre MUZY, Patricia Reynaud-Bouret
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 1 October 2022
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly

Description of the thesis topic

Spiking neural networks: learning and application to cognitive experiments

Classical Artificial Neural Networks (ANNs), especially those used in Deep Learning, are already inspired by biological neural networks. Synapses are modeled by synaptic weights that weight the "inputs" of a neuron before being summed. A non-linear transformation then gives the "output" of the neuron which is communicated to the post-synaptic neurons. The synaptic weights are optimized so that the network is able to, for example, classify images (Le Cun et al., 2015).
Skills : Mathematics and Computer science; mathematical/computational modeling skills and experience in object-oriented programming language (e.g., Python) are required. Basic knowledge in Deep Learning, Reinforcement Learning and experience with deep learning libraries (e.g., PyTorch or Tensor Flow) and architectures (CNN, etc.) will be appreciated.

One of the most current bio-inspired attempts consists in moving from a classical ANN, to a so-called "spiking" network: Spiking Neural Network (SNN) (Tavanaei et al., 2019). Indeed, neurons in a brain do not communicate by "sending" real numbers but electrical action potentials. These action potentials are essentially identical and it is more or less commonly accepted by biologists that the relevant information that is transmitted comes from the discharge times of the action potentials, which is commonly referred to as the spike train. This object, discrete in essence, can be well studied in computer science and mathematics.

The objective of this thesis is multiple: (i) We first want to better understand and formalize mathematically and computationally SNNs, (ii) Based on this formalization we want to propose a new class of SNNs, (iii) We want to apply this new class of SNNs to cognitive experiments implemented by our neurobiological and psychological colleagues.

References
Y. LeCun, Y. Bengio et G. Hinton : Deep learning. Nature, 521(7553):436–444, 2015.
A. Tavanaei, M. Ghodrati, S. R. Kheradpisheh, T. Masquelier et A. Maida : Deep learning in spiking neural networks. Neural Networks, 111:47–63, 2019.

Work Context

NeuroMod Institute and I3S lab of Université Côte d'Azur are situated at the Sophia Antipolis technology park (Europe's leading technology park). JAD laboratory is situated at Nice Valrose park. In computer science and mathematics, Université Côte d'Azur is a leading place with one the four French Interdisciplinary Institute for AI (3IA).
Skills : : Mathematics and Computer science; mathematical/computational modeling skills and experience in object-oriented programming language (e.g., Python) are required. Basic knowledge in Deep Learning, Reinforcement Learning and experience with deep learning libraries (e.g., PyTorch or Tensor Flow) and architectures (CNN, etc.) will be appreciated.

Constraints and risks

None

Additional Information

Interdisciplinary PhD. in Computer Science-Mathematics-Neuroscience :
- 06: Information sciences: foundations of computer science, computations, algorithms, representations, exploitations (I3S laboratory),
- 41: Mathematics and interactions (JAD laboratory),
- 51: Mathematical, computer and physical modeling for life sciences (NeuroMod Institute). "

We talk about it on Twitter!