Arnaud Doucet EE Engineering University of Melbourne Parkville, Victoria 3052 Australia [email protected] In this paper, we extend the Rao-Blackwellised particle filtering method to more complex hybrid models consisting of Gaussian latent variables and discrete observations. This is accomplished by augmenting the models with artificial variables that enable us to apply Rao-Blackwellisation. Ot...

In this paper, we focus on the statistical filtering problem in linear and Gaussian Jump Markov State Space Systems (JMSS). In such models, the computation of the optimal Bayesian estimate (in the sense of the mean square error) is an NP hard problem. Suboptimal solutions include algorithms based on numerical approximations or based on Sequential Monte Carlo methods. We propose an alternative o...

Sequential pattern mining algorithms using a vertical representation are the most efficient for mining sequential patterns in dense or long sequences, and have excellent overall performance. The vertical representation allows generating patterns and calculating their supports without performing costly database scans. However, a crucial performance bottleneck of vertical algorithms is that they ...

Inverse problems in geophysics require the introduction of complex a priori information and are solved using computationally expensive Monte Carlo techniques where large portions of the model space are explored . The geostatistical method allows for fast integration of complex a priori information in the form of covariance functions and training images. We combine geostatistical methods and inv...

We study the profit-maximization problem of a monopolistic market-maker. The sequential decision problem is hard because the state space is a function. We demonstrate that the belief state is well approximated by a Gaussian distribution. We prove a key monotonicity property of the Gaussian state update which makes the problem tractable. The algorithm leads to a surprising insight: an optimal mo...

Linear inverse Gaussian problems is traditionally solved using least squares based inversion. The center of the posterior Gaussian probability distribution is often chosen as the solution to such problems, while the solution is in fact the posterior Gaussian probability distribution itself. We present an algorithm, based on direct sequential simulation, which can be used to efficiently draw sam...

This paper develops a simulation-based approach to sequential parameter learning and filtering in general state-space models. Our methodology is based on a rolling-window Markov chain Monte Carlo (MCMC) approach and can be easily implemented by modifying state-space smoothing algorithms. Furthermore, the filter avoids the degeneracies that hinder particle filters and is robust to outliers. We i...

Biography Dr. Dan Cornford is a lecture in Computer Science and works in the Neural Computing Research Group at Aston University. Research interests are in the field of spatial statistics, space-time modelling and data assimilation. Lehel Csato is a post-doc in the same group working on an EPSRC grant (GR/R61857/01) looking at applying sparse sequential Gaussian processes to data assimilation. ...

Seizure events in newborns change in frequency, morphology, and propagation. This contextual information is explored at the classifier level in the proposed patient-independent neonatal seizure detection system. The system is based on the combination of a static and a sequential SVM classifier. A Gaussian dynamic time warping based kernel is used in the sequential classifier. The system is vali...