between Imperial College, Oxford, Bath, LSE, York and Edinburgh on “Network Stochastic Processes and Time Series”, funded by a multimillion pound EPSRC programme grant and several industry and government partners.
Learn morebetween Imperial College, Oxford, Bath, LSE, York and Edinburgh on “Network Stochastic Processes and Time Series”, funded by a multimillion pound EPSRC programme grant and several industry and government partners.
Learn morebetween Imperial College, Oxford, Bath, LSE, York and Edinburgh on “Network Stochastic Processes and Time Series”, funded by a multimillion pound EPSRC programme grant and several industry and government partners.
Learn moreThis 6-year programme (2022-2028) will bring probabilists, statisticians and data scientists together to study large dynamic networks with applications in medicine, transport, cybersecurity, environmental protection, finance, biology and economics.
We are delighted to announce the first of our NeST Online Seminar Series: Professor Anru Zhang (Duke University) on "High-order Singular Value Decomposition in Tensor Analysis".
learn moreAn invited papers session on Network Stochastic Processes and Time Series has been accepted for JSM2024. Speakers are Professors Mihai Cucuringu (University of Oxford), Tracy Ke (Harvard University) and Carey Priebe (Johns Hopkins University). Precise timing to be announced.
learn moreRSS 2024: Invited Session on Network Stochastic Processes and Time Series
learn moreThe 2024 Workshop on Statistical Network Analysis and Beyond (SNAB2024) is scheduled to take place on June 14-16, 2023 at the Courtyard by Marriott Nassau Downtown/Junkanoo Beach. Over the span of three days, this workshop aims to unite researchers in the field of network science and related disciplines, providing an avenue for the exchange of innovative ideas and recent findings. The workshop will encompass a wide range of topics, ranging from statistical network modeling to more extensive fields such as tensor modeling, deep learning, and text analysis.
learn moreWe are pleased to announce the successful conclusion of NeST’s first Away Day and Annual Meeting, which took place on 3rd-4th October in York. The two-day event featured a rich program of individual and snapshot presentations delivered by leading mid
learn moreOur first ISOC meeting took place on Wednesday, 24 May. It was a productive and engaging session, marked by valuable insights and thoughtful discussions. During the meeting, we had the privilege of having esteemed individuals from diverse backgrounds
learn moreThe first joint event took place online on the 21st of April 2023. There were 70 registered attendees (PhD candidates, Research Assistants, established researchers) from the following institutions: LSE, The Alan Turing Institute, Max Planck Institute
learn moreWe also aim to provide the community with curated network datasets and freely available software for practitioners and scientists wishing to analyse network data.
Network point processes often exhibit latent structure that govern the behaviour of the sub-processes. It is not always reasonable to assume that this latent structure is static, and detecting when and how this driving structure changes is often of interest. In this paper, we introduce a novel online methodology for detecting changes within the latent structure of a network point process. We focus on block-homogeneous Poisson processes, where latent node memberships determine the rates of the edge processes. We propose a scalable variational procedure which can be applied on large networks in an online fashion via a Bayesian forgetting factor applied to sequential variational approximations to the posterior distribution. The proposed framework is tested on simulated and real-world data, and it rapidly and accurately detects changes to the latent edge process rates, and to the latent node group memberships, both in an online manner. In particular, in an application on the Santander Cycles bike-sharing network in central London, we detect changes within the network related to holiday periods and lockdown restrictions between 2019 and 2020.
Learn moreData collected over networks can be modelled as noisy observations of an unknown function over the nodes of a graph or network structure, fully described by its nodes and their connections, the edges. In this context, function estimation has been proposed in the literature and typically makes use of the network topology such as relative node arrangement, often using given or artificially constructed node Euclidean coordinates. However, networks that arise in fields such as hydrology (for example, river networks) present features that challenge these established modelling setups since the target function may naturally live on edges (e.g., river flow) and/or the node-oriented modelling uses noisy edge data as weights. This work tackles these challenges and develops a novel lifting scheme along with its associated (second) generation wavelets that permit data decomposition across the network edges. The transform, which we refer to under the acronym LG-LOCAAT, makes use of a line graph construction that first maps the data in the line graph domain. We thoroughly investigate the proposed algorithm's properties and illustrate its performance versus existing methodologies. We conclude with an application pertaining to hydrology that involves the denoising of a water quality index over the England river network, backed up by a simulation study for a river flow dataset.
Learn moreWhen analyzing weighted networks using spectral embedding, a judicious transformation of the edge weights may produce better results. To formalize this idea, we consider the asymptotic behavior of spectral embedding for different edge-weight representations, under a generic low rank model. We measure the quality of different embeddings—which can be on entirely different scales—by how easy it is to distinguish communities, in an information-theoretical sense. For common types of weighted graphs, such as count networks or p-value networks, we find that transformations such as tempering or thresholding can be highly beneficial, both in theory and in practice.
Learn moreIn this work, we propose an adaptive wavelet-based approach for extracting primary dynamics in multivariate nonstationary time series.
Learn moreThe original generalized network autoregressive models are poor for modelling count data as they are based on the additive and constant noise assumptions, which is usually inappropriate for count data. We introduce two new models (GNARI and NGNAR) for count network time series by adapting and extending existing count-valued time series models. We present results on the statistical and asymptotic properties of our new models and their estimates obtained by conditional least squares and maximum likelihood. We conduct two simulation studies that verify successful parameter estimation for both models and conduct a further study that shows, for negative network parameters, that our NGNAR model outperforms existing models and our other GNARI model in terms of predictive performance. We model a network time series constructed from COVID-positive counts for counties in New York State during 2020--22 and show that our new models perform considerably better than existing methods for this problem.
Learn moreApplications are invited for Chapman Fellowships in the Department of Mathematics, commencing September 2024. There are several positions available, which will each be fixed term for 2 years. These posts provide an excellent opportunity for those seeking to pursue an academic career in Mathematics.
Learn MoreApplications are invited for a Research Officer in Statistics in the Department of Statistics at the London School of Economics. Commencing in September 2024, this is a fixed term appointment for 3 years. Closing date: 12 January 2024 (23.59 UK time); Interviews: 24 January 2024.
Learn MoreApplications are invited for a Postdoctoral Research Associate in Statistics in the Department of Statistics at the University of Oxford. Commencing in September 2024, this is a fixed term appointment for 3 years. Closing date: 12 January 2024 (12.00 noon, UK time); Interviews: 24 January 2024.
Learn moreWe invite applications for a Postdoctoral Research Associate to work on the EPSRC-funded programme grant ‘Network Stochastic Processes and Time Series (NeST)’ . NeST brings together the Universities of Bath, Edinburgh, Imperial College London, the London School of Economics and Political Science, Oxford and York, with industrial and government partners. This configuration has advantages for postdoctoral researchers as you will not only have access to your project leaders and team, but ultimately to all project members across the six institutions and the flow of ideas and problems across NeST. Stochastic network data are of rapidly increasing ubiquity in many fields such as medicine, transportation, cybersecurity, the environment, finance, biology and economics, and NeST aims to achieve a step change in the modelling and prediction of evolving, inter-connected stochastic network processes. As part of the NeST team, you will contribute to realising a substantial coordinated push to create, develop and apply innovative new models, computational techniques and underpinning theory, in response to real applied problems spurred by dynamic networks in many contexts. The research at the University of York will be led by Prof. Marina Knight and focus on the modelling and prediction of dynamically collected data at the nodes and edges of e.g. biological networks. The statistical challenge in such contexts is to more accurately reflect data characteristics in network time series models, where the data may be high-dimensional and for example exhibit nonstationarity, long-range dependence, and/or be driven by external factors. The position is full-time and will be held for 36 months, starting as soon as possible and ideally before 2 September 2024. Closing date: 13 March 2024; Interviews: 25 March 2024
Learn more