We're running three courses on state-of-the art aspects of mathematical and statistical modelling. They're aimed at postgraduate and post-doctoral environmental researchers.
An exciting feature of the course is that we encourage participants to bring their own project material as the basis for model formulation and coding sessions in small groups. Many students have left past courses with completely new working models for their projects.
Course leader: Prof Chris Roberston
Location: Each four-day course is fully residential at the University of Strathclyde in Glasgow.
Funding: There are 20 fully-funded places available on each course. This covers fees, travel, accommodation and subsistence. Priority is given to applicants whose research is at least 50% supported by NERC.
This course is based upon the premise that students have already taken a course of elementary statistical methods. The course assumes you have existing familiarity with programming in R, and of basic statistical concepts including probability and Bayes’ Theorem
Using examples from environmental science throughout, Bayesian model fitting methods will be introduced and compared to classical approaches. We will demonstrate how to set up complex structural models where Bayesian methods are necessary. Extensions to spatial and temporal smoothing, for example in disease mapping, will also be covered. The OpenBUGS statistical package will be used in conjunction with R for practical sessions. The focus will be on applications and modelling, particularly in areas cognisant with the students’ research, but will also cover sufficient theory to explain the modelling concepts
On completion of this course students will have a good understanding of the role of Bayesian methods in statistical modelling and will be aware of the advantages and disadvantages of this modelling in relation to traditional statistical modelling. They will also go away with working computer code which they can modify and reuse in their own work. They will also be able to critically review and assess Bayesian modelling reported in the environmental science literature.
Course Leader: Dr Douglas Speirs
Location: Each four-day course is fully residential based at the Ross Priory Conference Centrelocated on the beautiful shores of Loch Lomond.Funding: There are 9 fully-funded places available on each course. This covers fees, travel, accommodation and subsistence. Priority is given to applicants whose research is at least 50% supported by NERC.
This course will provide a basic understanding of modelling and it uses in environmental. The course will ‘lift the lid’ of mathematical models and teach you how to build your own ordinary differential equation models from the ground-up using the open-source R programming environment.
In this course we'll show you how to formulate biological, ecological, or physical problem in terms of ordinary differential equations, code and solve them in R. We'll start by examining simple food chain system and disease models, and progress to formulating and constructing basic models from scratch.
The course is designed to be very practically orientated, with short lecture sessions interspersed with hands-on practical work. With a high staff to student ratio, we can offer near 1-to-1 attention. In addition, we encourage you to come prepared to give a short presentation on your own projects. As a class we then consider how each of these projects can be posed in terms of differential equations, and work in groups to translate them into operation R-code. In the past, students have left with working models of their systems. This is the most innovative, exciting and challenging part of the course - especially for the tutors! We do not know in any detail in advance what problems the students will turn up with.
The course assumes existing familiarity with programming in R, and of basic calculus.