We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Department of Mathematical Sciences: Statistics & Probability Group


Statistics Seminars

Seminars are usually held on Monday at 14:00 in CM221, but you should check the seminar details for exceptions. Contact for more information about this seminar series.

Identifying the effect of public holidays on daily demand for gas

Presented by Sarah Heaps, Newcastle University

22 January 2018 14:00 in CM221

Gas distribution networks need to ensure the supply and demand for gas are balanced at all times. In practice, this is supported by a number of forecasting exercises which, if performed accurately, can substantially lower operational costs, for example through more informed preparation for severe winters. Amongst domestic and commercial customers, the demand for gas is strongly related to the weather and patterns of life and work. In regard to the latter, public holidays have a pronounced effect, which often extends into neighbouring days. In the literature, the days over which this protracted effect is felt are typically pre-specified as fixed windows around each public holiday. This approach fails to allow for any uncertainty surrounding the existence, duration and location of the protracted holiday effects. We introduce a novel model for daily gas demand which does not fix the days on which the proximity effect is felt. Our approach is based on a four-state, non-homogeneous hidden Markov model with cyclic dynamics. In this model the classification of days as public holidays is observed, but the assignment of days as ``pre-holiday'', ``post-holiday'' or ``normal'' is unknown. Explanatory variables recording the number of days to the preceding and succeeding public holidays guide the evolution of the hidden states and allow smooth transitions between normal and holiday periods. To allow for temporal autocorrelation, we model the logarithm of gas demand at multiple locations, conditional on the states, using a first-order vector autoregression (VAR(1)). We take a Bayesian approach to inference and consider briefly the problem of specifying a prior distribution for the autoregressive coefficient matrix of a VAR(1) process which is constrained to lie in the stationary region. We summarise the results of an application to data from Northern Gas Networks (NGN), the regional network serving the North of England, a preliminary version of which is already being used by NGN in its annual medium-term forecasting exercise.

Contact for more information

Presented by Christina Goldschmidt, University of Oxford

29 January 2018 14:00 in CM221

Contact for more information

Extensive condensation in preferential attachment with choice

Presented by Nic Freeman, Sheffield University

5 February 2018 14:00 in CM221

I will introduce a new preferential attachment model in which each vertex has an associated fitness value. I will discuss the behaviour of the model as the number of nodes tends to infinity, including the existence of a "condensation" phase in which a small number of especially fit vertices are able to (temporarily) gain disproportionately large degrees. The work relies on a new connection between preferential attachment with fitnesses, and branching-coalescing particle systems; it leads to a clear and simple explanation for why the condensation phenomenon occurs.

Contact for more information

Presented by Theodore Kypraios, University of Nottingham

12 February 2018 14:00 in CM221

Contact for more information

Presented by Peter Nejjar, IST Austria

19 February 2018 14:00 in CM221

Contact for more information

ABC for expensive simulators

Presented by Richard Everitt, University of reading

26 February 2018 14:00 in CM221

Approximate Bayesian computation (ABC) is now an established technique for statistical inference in the form of a simulator, and approximates the likelihood at a parameter θ by simulating auxiliary data sets x and evaluating the distance of x from the true data y. Synthetic likelihood is a related approach that uses simulated auxiliary data sets to contract a Gaussian approximation to the likelihood. However, these approaches are not computationally feasible in cases where using the simulator for each θ is very expensive. This talk investigates two alternative strategies for inference in such a situation. The first is delayed acceptance ABC-SMC (, in which a cheap simulator is used to rule out parts of the parameter space that are not worth exploring. The second is bootstrapped synthetic likelihood (, which uses the bootstrap to cheaply estimate the synthetic likelihood. We also examine a synthetic likelihood approximation that is constructed, using the bag of little bootstraps, from subsampled data sets. Applications to stochastic differential equation models and doubly intractable distributions will be presented.

Contact for more information

Presented by Jonathan Rougier, University of Bristol

12 March 2018 14:00 in CM221

Royal Statistical Society North Eastern Local Group meeting

Contact for more information

Other seminar series

RSS North East Local Group seminars

The North Eastern Local Group of the Royal Statistical Society organises an annual programme of meetings and events allowing statisticians across the North East of England to meet and discuss topics of interest. The meetings are free to attend and non-members are always welcome. Meetings are typically held in Durham or Newcastle.

The program of these regular meetings can be found at the Royal Statistical Society North Eastern Local Group home page.

Postgraduate Seminars

These seminars offer an opportunity to find out more about what other postgraduates in the department are studying and help the speakers to improve their presentation skills in an informal atmosphere. Occasionally, postgraduate seminars are given by a member of staff. The postgraduate seminar will start in Epiphany term 2014 under the title "Stats4Grads". Postgraduates from the Statistics and Probability group, as well as from other groups and Departments, are more than welcome to attend, and to present their work with quantitative focus to a postgraduate audience.

Please see details at the Stats4Grads webpage.

Postgraduate Training Weeks

The Statistics group runs a series of postgraduate training weeks jointly organized with Newcastle University, which include lectures, seminar talks by guest speakers, computer practicals, and also seminar talks by postgraduate students. Six different training courses are offered in a 3-year-cycle, covering Multivariate Distributions (Newcastle), Smoothing (Durham), Statistical Computing (Newcastle), Foundations of Statistics (Durham), Design (Newcastle), and Modelling (Durham).