About the Project
An enormous variety of complex systems shares the unifying property that they can be mathematically modelled as a network of interacting elements. Examples of this include social iterations, communication systems, cell metabolism, transportation infrastructure, among many others. Despite the different domains, all these systems can be modelled at their most fundamental level under the same network formalism. With the aim of exploiting this universality, a great deal of transdisciplinary research has been devoted to developing general network models that are valid across different domains.
The aim of this PhD project is to move towards this goal using a specific blend of mathematical modelling and data analysis, based heavily on concepts and analytical tools from Statistical Physics, and employing a variety of approaches from Bayesian Statistical Inference and Machine Learning. In particular, the main objectives are:
1. Elaboration of generative models of networks that take into account key evolutionary aspects (e.g. optimization towards robustness under constraints, homophily, incremental growth dynamics), and yield credible descriptors of large-scale network structure (e.g. modular organisation, hierarchies and centralisation).
2. Development of principled inference methods that can extract model parameters from real-world network data, as well as model selection approaches that can identify the most appropriate generative process based on empirical evidence.
3. Employment of the modelling and inference frameworks to make predictions that generalise from past observations, identify errors and omissions in data, as well as opportunities for architectural improvements.
The combination of these three objectives would yield concrete connections between the structure, function and evolution of network systems, with potential applications as diverse as preventing the outbreak of diseases and traffic jams, discovering new interactions between drugs, and building a censorship-free internet.
Furthermore, the diverse and multidisciplinary nature of the research would give the candidate many options in further pursuing an academic career in Theoretical Physics, Machine Learning and Data Science, as well as opportunities for applications in industry.
The successful candidate should be highly motivated and have a degree in Physics, Applied Mathematics or related fields. Demonstrable familiarity with mathematical modelling as well as computational skills (C/C++ and/or Python) is essential.
The position is for 3.5 years of full-time study and will administratively belong to the Department of Mathematical Sciences at the University of Bath, associated with the Centre for Networks and Collective Behaviour, and will be supervised by Dr. Tiago Peixoto ([Email Address Removed], https://skewed.de/tiago).
The successful candidate will be ready to start by March 2018 at the latest.
Applications may close early if a suitable candidate is found; therefore, early application is recommended.
Funding Notes
UK and EU students applying for this project may be considered for a University Research Studenship which will cover Home/EU tuition fees, a training support fee of £1,000 per annum and a tax-free maintenance allowance of £14,553 (2017/18 rate) for 3.5 years.
Note: ONLY UK and EU applicants are eligible for the studentship; unfortunately, applicants who are classed as Overseas for fee paying purposes are NOT eligible for funding.
Continue with Facebook
