arrow drop search cross

Parameter Learning for Spiking Neural Networks Modelled as Timed Automata and Epidemiological Models for Solving a Water Distribution Problem

Seminar held on September 12, at 10:00, in the MSI's premises in Sophia Antipolis

12/09/2019   :   10h00
Salle des séminaires MSI Sophia Antipolis
 Speakers: Elisabetta De Maria and Enrico Formenti, I3S Sophia Antipolis
Publication : 12/09/2019
Partager cet article :

Speakers: Elisabetta De Maria and Enrico Formenti (I3S Laboratory, Sophia Antipolis)

Abstract: In the first part of the talk, given by Elisabetta De Maria, we introduce a new approach to learn the synaptical weights of neural biological networks. At this aim, we consider networks of Leaky Integrate and Fire neurons and model them as timed automata networks. Such a formal encoding is exploited to find an assignment for the synaptical weights of neural networks such that they can reproduce a given behavior.

In the second part of the talk, given by Enrico Formenti, we focus on solving a water distribution problem. It is well-known that water supply is becoming a global issue both at the level of drinking water and the watering for agriculture, industry, etc. We show how to use an epidemiological model based on neural networks for outbreak detection for optimizing watering in agricultural applications.