I’m trying to start and finish a project in the next month. The project involves a lot of thinking and programming. I’m trying to solve a large action space Factored MDP. It’s challenging because the actions can’t be factored. It’s almost a flat MDP.
I’m hoping that the results will be easily interpretable. There is no point finding an optimal solution if I can’t explain it :-)
I’ll try to report the progress I make as we go.
This year, our team has been working on a cost benefit analysis project to determine the priority threat management options to protect the listed species of the Pilbara, WA.
Part of our project consists in gathering data from different sources, published and grey literature, and most importantly experts knowledge. We ran an anonymous expert elicitation workshop in March 2013 to gather the critical information needed for our project.
When running an expert elicitation exercise it is recommended that experts remain anonymous so that bias is limited when collecting the expert knowledge (benefit estimates, in this case). Once the preliminary analysis is done and individual mistakes are corrected, it is important to provide a way of discussing the results as a group so that additional information can be exchanged and informed discussion can lead to less divergence in opinions.
To set up that discussion phase and maintain an anonymous process, we have set up a forum for our experts to comment and discuss the current group trends and divergence. We’ve had a couple of responses to our forum from contributors who said that they’d appreciate more information about how to use the forum. In response I recorded a quick video that captures my screen as I go through the features of the forum. The video is available on YouTube, and can be accessed at the link below. You can play this video full screen and you can also increase the quality to High Definition (HD).
We recently had a paper accepted in Methods in Ecology and Evolution. I can’t say too much about it yet, but I hope it will be well received and useful to many. Well done Lucile and team!
Marescot L., Chapron G., Chadès I., Fackler P.L., Duchamp C., Marboutin E. & Gimenez, O. (Accepted 10/06/2013) Complex decisions made simple: a primer on stochastic dynamic programming. Methods in Ecology and Evolution.