The QMEE CDT Project proposal database

Welcome to the QMEE CDT Project proposal database. This is a live list of projects proposals put forward by PIs across the CDT partner institutions

PIs/Supervisors will continue to add projects to this list over the next few months, so do keep checking back! You can search the projects using the box below: simply enter some text and press Search to do a text search across all the database fields. If you want to search more finely, the search tool also allows you to search on particular details of the project descriptions: you will see these finer search options appear if you click on the search box.

Click on the view button next to a project to get the full proposal description. If you want to download project details, either for all projects, or for a subset you have searched for, then click on the 'Download details' button.

Metric and Kernel Inference for Coupled Ecological and Social Systems
A generic environmental challenge is to influence ecological systems and to influence the ecological beliefs of human social systems. Ecological systems can often be mapped to elements which can have binary states linked via a coupling function (or kernel) [Nobel]. So too social systems can often be thought of as the binary opinions of individuals about e.g. climate change/recycling that are coupled together [Castellano,Egan,Howe]. Statisticians call such systems Binary Markov Random Fields and Physicists call them spin-systems/Ising-models. We might have information about the state of each element (e.g. a tree’s yield in a grove or an individual’s climate views in a society) and further know their co-ordinates in some space (e.g. GPS co-ordinates, local temperature, age). Those co-ordinates define a space in which, if we have a metric, we can measure distances between the different elements such that the interaction between the elements decays with distance. Unfortunately, we often lack an idea of the relevant metric and coupling function. If we don't know how to weight e.g. a separation of 1m in physical distance relative to e.g. a separation of 1oC, then we need to infer either a metric on the space or a kernel function that tells us how the elements of the system couple together. To add insult to injury the state of the system itself could be modulated by its local co-ordinate (e.g. a tree’s state might depend on its local temperature and how it couples to another system might depend on how different their local temperatures are). Our question is thus "given observations of only each element’s state and their co-ordinates in a space can we infer both a metric and whether the local co-ordinates are also influencing element state". The next question is "if we can only observe a small proportion of the total systems how large must that proportion be in order to do good inference?" The student will draw on empirical data from [Noble] from new survey data on climate and ecological opinions and to simulated data. With good inference of metrics and kernels we can then move towards robust techniques to influence crops and ecosystems and to influence opinions about climate change. Emergent long-range synchronization of oscillating ecological populations without external forcing described by Ising universality, Noble, Machata, Hastings, Nature Communications (2015). Recent improvement and projected worsening of weather in the United States, Egan & Mullin, Nature (2016). Global perceptions of local temperature change, Howe, Markowitz, Lee, Ko, Leiserowitz, Nature Climate Change (2013). Statistical Physics of Social Dynamics, Castellano, Fortunato, Loreto, Reviews of Modern Physics (2009).
Nick Jones
Richard Everitt
Development of mathematical theory, Computing
Nick Jones
This project principally involves the development of new inference techniques. The student will also be able to develop small social surveys. The requisite skill set will be covered in undergraduate/masters degrees in mathematics/statistics/physics but might be covered in other courses also.
There has been very little work on optimal metric inference for general coupling functions (especially when the elements lie in spaces defined by categorical, ordinal and continuous co-ordinates). Despite this there is great scope for the ecological and sociological application of the methods developed.
The exploration of spin-system models of ecological systems is a vibrant space -- our work will widen the applicability of models of this kind. We will be able to perform model selection and thus identify whether e.g. conjectured co-ordinates of relevance for the coupling of ecological systems, like local temperatures, are of practical relevance.
If we can identify the correct models for ecological and social systems then we can move towards optimally controlling those social and ecological systems. This can influence crop control strategies but also public information campaigns linked to climate and ecological awareness.
Networks on N nodes need an amount of data collection scaling like O(N^2) whereas finding the co-ordinates of nodes needs O(N). If our inference techniques succeed we can allow the identification of network ensembles with O(N) amount of effort (the typical amount used in surveys): allowing a huge increase in the applicability of network-science ideas.
This project mixes ecology, statistics, and sociophysics for the study of ecological systems and climate opinions.
Climate and climate change, Population ecology, Ecological/Evolutionary tools, technology & methods
The student will be trained in topics in network science, sociophysics and ecological models (Nick Jones) and statistical inference for spin systems (Nick Jones and Richard Everitt). We have a track record of our students doing internships in Data Science and in spinning out Data Science firms.
Imperial College Mathematics and Reading Mathematics
No
2017-10-02 11:40:06