Conference Papers
http://hdl.handle.net/10453/14
2014-04-24T19:46:18ZSolution of Extreme Transcendental Differential Equations
http://hdl.handle.net/10453/26441
Solution of Extreme Transcendental Differential Equations
Nettleton, Stuart
Blind peer review
Extreme transcendental differential equations are found in many applications including geophysical climate change models. Solution of these systems in continuous time has only been feasible with the recent development of Chebyshev solvers such as Mathematica 9?s NDSolve function. This paper presents the challenges and means of solving the widely used DICE 2007 integrated assessment model in continuous time. Application of the solution technique in a mobile policy tool is discussed.
2013-01-01T00:00:00ZCorrLog: Correlated Logistic Models for Joint Prediction of Multiple Labels
http://hdl.handle.net/10453/23498
CorrLog: Correlated Logistic Models for Joint Prediction of Multiple Labels
Bian, Wei
Lawrence, Neil; Girolami, Mark
In this paper, we present a simple but effective method for multi-label classification (MLC), termed Correlated Logistic Models (Corrlog), which extends multiple Independent Logistic Regressions (ILRs) by modeling the pairwise correlation between labels. Algorithmically, we propose an efficient method for learning parameters of Corrlog, which is based on regularized maximum pseudolikelihood estimation and has a linear computational complexity with respect to the number of labels. Theoretically, we show that Corrlog enjoys a satisfying generalization bound which is independent of the number of labels. The effectiveness of Corrlog on modeling label correlations is illustrated by a toy example, and further experiments on real data show that Corrlog achieves competitive performance compared with popular MLC algorithms.
2012-01-01T00:00:00ZGeneralised Extreme Value geoadditive model analysis via variational Bayes
http://hdl.handle.net/10453/19063
Generalised Extreme Value geoadditive model analysis via variational Bayes
Nevillea, Sarah; Wand, Matthew
Alfred Stein, Edzer Pebesma and Gerard Heuvelink
We devise a variationalBayes algorithm for fast approximate inference in Bayesian GeneralizedExtremeValue additive modelanalysis. Such models are useful for flexibly assessing the impact of continuous predictor variables on sample extremes. The new methodology allows large Bayesian models to be fitted and assessed without the significant computing costs of Monte Carlo methods
2011-01-01T00:00:00ZQuasi-Monte Carlo methods for derivatives on realised variance of an index under the benchmark approach
http://hdl.handle.net/10453/19062
Quasi-Monte Carlo methods for derivatives on realised variance of an index under the benchmark approach
Baldeaux, Jan; Chan, Leung Lung; Platen, Eckhard
William McLean and Anthony John Roberts
We apply quasi-Monte Carlo methods to the pricing of derivatives on realised variance of an index under the benchmark approach. The resulting integration problem is shown to depend on the joint density of the realised variance of the index and t he terminal value of the index. Employing a transformation mapping for this joint density to the unit square reduces the difficulty of the resulting integration problem. The quasi-Monte Carlo methods compare favourably to Monte Carlo methods when applied to the given problem.
2011-01-01T00:00:00Z