Laurent Hébert-Dufresne
News
-
Postdoctoral Fellowship: It's so crazy, it's amazing!
Things are moving forward at an incredible speed. A lot of projects were started, and a few were completed, during my 5 weeks stay at the Santa Fe Institute. I can’t wait to put some of those projects online. I’ve also been selected by the James S. McDonnell Foundation for their Postdoctoral Fellowship Awards in Studying Complex Systems . I’m not sure where I’ll have the pleasure of continuing my research, “but I have a feeling it will be a place both wonderful and strange”.
-
Currently in Santa Fe, but things are shaking in Québec
Currently at the Santa Fe Institute for a one month visit which is already very interesting. But in Québec, the Dynamica group has uploaded a new paper on the arXiv. In Coexistence of phases and the observability of random graphs , we use a version of Antoine’s formalism for bond percolation on multitype graphs to investigate “depth-L percolation”, where observing one node implies observing his neighbours up to a distance L. Interestingly, extensive components can coexist! Meaning that monitoring a macroscopic portion of a graph does not prevent a macroscopic event to occur unbeknown to the observer . In your face NSA!
-
New paper (finally) out
Our new paper, Global efficiency of local immunization on complex networks , was published today in Nature’s open access (!) journal: Scientific Reports. This is a side project that started almost two years ago with Antoine Allard and Jean-Gabriel Young. This paper had a torturous development, but thanks to The Long Road , is now much more complete. In fact, there’s a new (nice) result on a change in the “characteristic scale” of epidemic dynamics under change of transmissibility.
-
The beginning of the end (part 3)
Another paper of my CSSS12 group was published, again on antiviral resistance but without networks this time. Head over to the PLOS ONE website to have a look at Oscar’s paper on Optimizing Treatment Regimes to Hinder Antiviral Resistance in Influenza across Time Scales . Most interesting result (for me): a treatment regime minimizing risk of resistance emergence is obtained analytically without considering any network structure, yet the result is valid for simulations on heterogeneous networks (even though the epidemic threshold of the resistant strain depends highly on this heterogeneity of the contact structure).
-
Road trip and NetSci 2013
I’ll be AFK from mid-May for a road trip in Iceland, but will reconnect as soon as we get to Copenhagen for NetSci 2013 . I’ll be giving a talk on a small project called Hard-core Random Networks on a model of random networks which preserves a given shell decomposition structure, maps to a weird version of the Configuration Model and is a quite efficient proxy for analytical predictions of bond percolation. Antoine Allard will be presenting his much broader analytical formalism for bond percolation on any kind of random networks (including the hard-core ones) and Jean-Gabriel Young will present part of his on-going work on community detection (different from the COMPLEX 2012 talk).