Presenting at ICTACS 2010

I will present in 2 days at the ITACS'10 conference in Can Tho. Actually, i'm already in this nice and quiet city since last saturday. I took part in a the IFI's summer school.

This summer school was about "Multi-agents systems from the complex system perspective", which is quite my speciality ;-), and "Formal validation techniques of communicating systems" which was quite interesting even though it's not my field.

Anyway, the ICTACS conference is (quite) general conference about "computer science" nevertheless it's still an international conference which will be published (locally).

These two events take place at the Can Tho university which seems to be really dynamic and growing fast. Actually, they got a computer team, called 'DREAM', which just joined a support program from IRD.

You can check my presentation slides: here.

GAMA 1.3 is out

We recently took part in a summer school about epidemiological modeling which took place at Gaston Berger university, St Louis, Senegal. I will write about it later on...

Anyway, we needed a stable and up-to-date release of GAMA for the teaching. This has been done in the beginning of July and it is now available on the website. In addition, documentation have been updated and 2 new "epidemiological" tutorials have been created (french only for the moment).

Even though, the 1.4 version of GAMA is due quite soon, end of September, do not hesitate to test the 1.3 one as it includes a lot of "easying" GIS functions and improved general stability and performance!

GAMA 1.3 demo

Today, I wanted to show a very brief demo of GAMA 1.3 alpha1. To do so, I run my current GAMAVI model (some explanation of this model are below):

The system represents a village of North Vietnam, the topography of the village is based on a google map image (see the background). Within the village you have ground (brown color), household (white), farms (very light pink) and ponds (blue). Whenever a environmental place is infected it becomes orange (more like yellow in the video). This village is surrounding by ricefield lots (green). The agents (circle) are poultry flocks. There are two types: free range ducks which go to the ricefields and "pond ducks" which go to the ponds, both during daytime. Each flock contains a matrix of individuals where their infectious state only is represented. Whenever a flock has one infected individuals it excretes one unit of virus in the current environment (but not in farm). Then the environment cell will deplete regularly its virus level. When a flock is localized in a infected environment it can trigger the infection of one individual.

This model is pretty simple compared to my objective model and it is using reasonable parameters but not the ones i will have finally.

By the way, as you might guess this model does not use a grid for environment but a continuous space containing geometries (polygons) read directly from GIS data (or generated from it in the case of rice-field lots).

GAMA 1.3 alpha1 is out!

The team behind the GAMA simulation platform, I use for simulated my models, as just released the first alpha version of GAMA 1.3. Although, only source code is available as it is an alpha only! More info on the GAMA google group.

In additon, the new Eclipse version has been released as well yesterday (GAMA is developed using this IDE).

Edit: We just discovered that GAMA 1.2 has been made available at Softpedia!

First review!

I have been an (unofficial) reviewer for the first time this week!

It is a paper for the AIMAS conference'10. I prefer not to tell what subject as review are anonymous (not even sure i should state the conference as I reviewed as a request from the original reviewer). Despite it was not my speciality I'm quite confident that the contribution of the paper is quite small and the quality of english is only acceptable.

Creation

As a first post on this blog i should state its origin. Indeed, it is quite short: i am a PhD candidate in agent-based modeling and i got the inspiration from gisagents.

I don't really know what i would focus on. Anyway, it will revolve around agent-based models, their use in epidemiology, epidemiological modeling in a more general way and probably some methodology on ABM especially communication of model ( especially the ODD protocol).