Workshop Announcement: Fitting Flexible State-Space and Hierarchical Models Using the Laplace Approximation and Automatic Differentiation

I, Mollie, am organizing a workshop here at the University of Zurich. I’ve recruited two experts to come help me teach: Hans Skaug (Dept of Mathematics, Univ of Bergen) and Kasper Kristensen (Dept of Applied Maths and Comp Sci, Technical Univ of Denmark). We’ll teach researchers from diverse fields how to fit state-space models via maximum likelihood estimation using the Laplace approximation to integrate out latent variables and automatic differentiation to calculate gradients. This is much faster than Bayesian methods.

These methods are implemented in AD Model Builder and a new R package called TMB. Currently, the course organizers are trying to decide if we should teach both, or only TMB. There are costs and benefits to being early adopters of new software. For example, several years ago, when the user base of ADMB was expanding beyond fisheries stock assessment, we found several bugs that weren’t exposed until people started trying different types of models and different computational platforms. We don’t want to expose students to these issues. On the other hand, learning one new program (rather than two) in a three day workshop might be preferable for some participants. Plus, most ecologists already use R, so learning the package TMB might be slightly easier for them than learning to use ADMB through the R2admb interface. TMB is more streamlined than ADMB because it had ADMB as an example and took advantage of existing free and open source C++ libraries. Whatever we decide to do, we’ll make sure the class exercises and examples are fully debugged before the workshop. Maybe I’ll use my lab mates’ computers as Guinea pigs.

The course will take place September 1-3, followed by 2 days of developing the TMB package and applications. We still have some openings for participants. Applications are due August 1.

More information can be found at the course website

The workshop is funded by a GRC Grant from the University of Zurich.

Mollie Attended an ADMB Developers’ Workshop in Iceland

After the ADMB Developers' Meeting, I investigated the thermal activity in Iceland.

This photo has nothing to do with the workshop. I did some hiking after the workshop and found this huge boiling cauldron of water in the land of fire and ice. The landscapes were amazing with so much geothermal activity, old lava fields, and water falls.

The main purpose of my trip to Iceland was to work September 18-22 with eleven other developers on the statistical software Automatic Differentiation Model Builder (ADMB) at the University of Iceland and the Marine Research Institute in Reykjavik, Iceland. ADMB is useful for fitting nonlinear models and has the flexibility to fit random effects. It’s nearly as flexible as MCMC methods like WinBUGS, but much faster and without the need for specifying prior information (it’s really cool!).

It’s so much easier to get everyone engaged in a discussion and making decisions when we’re all working in the same room, rather than over email when we’re spread out over the corners of the earth. Since we had developers coming from North America and Europe, Iceland was a good midway point, and we had an excellent Icelandic host.

We discussed many important decisions for how to move forward with the software including documentation, parallelization, the most efficient linear algebra libraries, optimizers, links to R, and potentially moving to Github. We’re working to make the documentation easier for new users and writing an introduction aimed especially at R users. We discussed teaching workshops in the near future. We have one scheduled at the International Statistical Ecology Conference and we’re interested in teaching to other groups as well. Let me know if you’re interested.

Mollie Brooks, our new quantitative ecologist

We have a new quantitative ecologist joining our ranks, Mollie Brooks, who is a recent PhD graduate from University of Florida, Gainesville. Fresh out of Ben Bolker’s group, Mollie is bringing with her much needed skills in statistical and demographic analysis, and to our pleasant surprise, in baking!

During her postdoc, Mollie will contribute to our research projects on early warning signals and resurrecting past eco-evolutionary responses.


Mollie Brooks | Postdoc


After her postdoc in Zürich, Mollie moved to Copenhagen to work as a biostatistician in fisheries research.

My main interest is in quantifying life history traits and their effects on demography. I think it’s interesting that tradeoffs are a ubiquitous part of life and the ways that species and individuals handle those tradeoffs affect their fitness. I have also worked on modelling community dynamics, epidemiology, and the evolution of sex-change. Most of my work involves either fitting models to data or writing equations. Tools and methods I use include R, TMB, maximum likelihood estimation, MCMC sampling, and dynamic optimization modelling.

2013-2016 Postdoctoral Research Associate, Institute of Evolutionary Biology and Environmental Studies, University of Zurich
2012 PhD in Biology, Department of Biology, University of Florida
2004 BS in Mathematics with related studies in Computer Science, University of Pittsburgh

Predicting population responses to environmental change

species2A major goal in biodiversity conservation is to predict responses of biological populations to environmental change. To achieve this goal, we must identify early warning signals of the demographic changes that underlie sudden population declines or explosions. Some studies have achieved phenomenological prediction of sudden changes, but recent advances that link trait-based information with demography hint that a mechanistic understanding is within reach. We are developing a predictive framework by investigating how wildlife populations respond demographically, ecologically and evolutionarily to environmental change, and identifying the demographic and phenotypic statistics that can be used as early warning signals of population change. This project will exploit nine unique mammalian systems to identify early warning signals of population change and test these signals on two experimental systems. The results will hopefully provide much-needed predictive insight into how wildlife populations respond to rapid environmental change.


In collaboration with: