22nd February 2012, 5:30pm
PhD Student talks at Lancaster University
Efficient detection of multiple changepoints within an oceanographic time series, Rebecca Killick (Lancaster University)
Abstract
We consider the problem of detecting multiple changepoints in large oceanographic data sets. In this setting the amount of data being collected is continually increasing and consequently the number of changepoints will also increase with time. An efficient and accurate analysis of such data is of considerable interest to those working in the energy sector as understanding the characteristics of the ocean environment is central to reliable design and operation of marine and coastal structures. Detecting the presence of changepoints in oceanographic time-series is of particular importance, since statistical and engineering modelling of the ocean environment, structural loading and response typically assumes stationarity of the environment (in time). Drawing on recent work on efficient search methods by Killick et al. (2011), we compare and contrast the effect of different approaches to this data, focusing in particular on computational and statistical aspects. The talk will conclude by highlighting the importance of such computationally efficient methods in an oceanographic setting.
Particle filters for target tracking, Chris Nemeth (Lancaster University)
Abstract
The need to analyse sequential data is a problem found in target tracking, where sensors relay the position of a target in real-time. The data received from the sensors are often noisy and require the noise from the data to be filtered, revealing the true position of the target. Traditional methods utilised for these problems include the Kalman filter which is an effective method of solving the filtering problem when the models of interest are linear-Gaussian. However, when there are nonlinearities in either the model or observation process, the Kalman filter no longer provides optimal filtered estimates. A popular alternative method for such scenarios, known as particle filtering, creates a discrete approximation of the posterior distribution of the target’s state. The approximation is then updated and evolves as new observations become available. In this talk I will introduce the basic concepts behind particle filtering and compare this approach to Kalman filtering using an example from the target tracking literature.
A study of the extreme risk of financial investments, Ye Liu (Lancaster University)
Abstract
Standard statistical inferences usually rely on a good amount of information, whereas the most influential financial events like the credit crunch have occurred only a few times throughout human history. Extreme value theory, as a brunch of statistics that deals specifically with rare events, provides a scientific tool for understanding the uncertainty when the financial market is in a critical condition. Our research is based on a recently developed multivariate extreme value model and its application in finance. In particular our interest is to quantify the extreme risk of a typical portfolio of financial investments.
Venue: Postgraduate Statistics Centre, A54 Lecture Theatre, Lancaster University
No charge is made to attend meetings and non-members are welcome.
