Big news about MAXENT

Ian Renner and I just published a paper in Biometrics show in the MAXENT is mathematically equivalent to a GLM (and hence related to poisson point process models).  See the Methods Blog for a more detailed post on this result and its implications.

Welcome Sara Taskinen

Welcome to Sara Taskinen of the University of Jyvaskyla, Finland (pictured below - she's the one in the middle!).  Sara will be visiting us here at UNSW for all of 2013.  She is an expert in robust statistics especially with multivariate applications, so we've been doing some work robustifying methods of estimation and inference about allometric lines, i.e. modifying how MA and SMA lines are estimated so that they are less sensitive to outliers.  Some of our earlier work has already been incorporated into the SMATR package, and a paper on some of our earlier ideas on how to construct robust confidence intervals for slope was recently published in the Biometrical Journal.  We've since been robustifying methods of hypothesis testing about one or several allometric lines - paper and code coming soon!

Eco-Stats symposium in Sydney Australia, July 11-12 2013

Congrats all round

Some good news at Eco-Stats in the last week or so:
- David was awarded an ARC Future Fellowship for 2012-2016, to focus his time on developing tools for predictive modelling of communities using multivariate data in ecology.
- Alice just landed a tenure-track position at Hong Kong Baptist University, starting September 2012.  Super news.
- Our visitor Alireza Faraz just landed a prestigious tenure-track position at U. Liege.
Congrats all round - party time!

Welcome Alireza

Dr Alireza Faraz, U. Liege Belgium, is visiting Eco-Stats until the end of the year.  He has expertise in the application of control charts to multivaraite data, and will focus on applying such methodology to multivariate abundance data in ecology.
Welcome Alireza!

Eco-Stats paper rated in Faculty of 1000

Our recent paper on confounding of mean and variance effects in multivariate analysis has recently been rated as a "Must Read" in the Faculty of 1000, a platform providing post-publication peer-review and selecting only the most important articles in biology and medicine. Just 2% of published articles are highlighted on Faculty of 1000 each month, and I'd say the rate is a lot lower when you get outside of medicine and biotech, so it's a bit of a coup getting an Eco-Stats paper in there!

The review was contributed by Ferdinando Boero and Stanislao Bevilacqua, Universita' del Salento, Italy, and highlights the importance of thinking critically about the properties of methods of data analysis, and how suitable they are to the data at hand.  We think this is a particular issue in multivariate analysis in ecology at the moment.

Welcome Jakub!

Welcome to Jakub Stoklosa, who has joined Eco-Stats fresh out of a PhD on various issues related to capture-recapture modelling, supervised by Prof Richard Huggins at U Melb.  Jakub will be working on fast model selection algorithms for correlated count data, and how to apply them to develop flexible new models for correlated counts.

To find out more about Jakub and his previous pursuits, check out his recent Biometrics article, or his department seminar coming up on May 4 at 4pm, OMB-145.