University of Birmingham

Birmingham, United Kingdom

Departments View all

School of Biosciences
8,572
Total Impact Points
264
Members
School of Psychology
4,389
Total Impact Points
230
Members
School of Geography, Earth and Environmental Sciences
1,111
Total Impact Points
196
Members

Publication History View all

  • [Show abstract] [Hide abstract]
    ABSTRACT: A Generalised Cornu Spiral (GCS) is a planar curve defined to have a monotonic rational linear curvature profile and as such these curves are considered fair. However, their implementation in current CAD systems is not straight forward partly due to not being in the usual polynomial form. A GCS cannot be expressed exactly using a finite polynomial and so a compromise can be achieved by instead approximating the GCS with a suitable polynomial. An efficient robust approximation of the GCS using quintic polynomials is presented. The approximation satisfies the G2G2 continuity conditions at the end points and the remaining four degrees of freedom are argued for by looking at G3G3 approximations. The method begins by reparameterising the GCS in terms of more intuitive geometric descriptions; the winding angle, change in curvature and a shape factor. The G3G3 approximations provide insight to help define values for the free parameters, and the new geometric form allows for the shortcomings in the G3G3 approximations to be controlled. The efficiency of the approximation is improved compared to earlier methods which required a numerical search. Also, there is strong evidence that the method guarantees a satisfactory approximation when the GCS lies within certain identified bounds.
    Journal of Computational and Applied Mathematics 01/2015; 273:1–12.
  • [Show abstract] [Hide abstract]
    ABSTRACT: The classical machinery of supervised learning machines relies on a correct set of training labels. Unfortunately, there is no guarantee that all of the labels are correct. Labelling errors are increasingly noticeable in today׳s classification tasks, as the scale and difficulty of these tasks increases so much that perfect label assignment becomes nearly impossible. Several algorithms have been proposed to alleviate the problem of which a robust Kernel Fisher Discriminant is a successful example. However, for classification, discriminative models are of primary interest, and rather curiously, the very few existing label-robust discriminative classifiers are limited to linear problems. In this paper, we build on the widely used and successful kernelising technique to introduce a label-noise robust Kernel Logistic Regression classifier. The main difficulty that we need to bypass is how to determine the model complexity parameters when no trusted validation set is available. We propose to adapt the Multiple Kernel Learning approach for this new purpose, together with a Bayesian regularisation scheme. Empirical results on 13 benchmark data sets and two real-world applications demonstrate the success of our approach.
    Pattern Recognition. 11/2014; 47(11):3641–3655.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Hydrogen production through fuel reforming can be used to improve IC (internal combustion) engines combustion characteristics and to lower vehicle emissions. In this study, a computational fluid dynamics (CFD) model based on a detailed kinetic mechanism was developed for exhaust gas reforming of biogas to synthetic gas (H2 and CO). In agreement with experimental data, the reactor's physical and chemical performance was investigated at various O2/CH4 ratios and gas hourly space velocities (GHSV). The numerical results imply that methane reforming reactions are strongly sensitive to O2/CH4 ratio and engine exhaust gas temperature. It was also found that increasing GHSV results in lower hydrogen yield; since dry and steam reforming reactions are relatively slow and are both dependent on the flow residence time. Furthermore, the hot spot effect, which is associated to oxidation reforming reactions, was investigated for catalyst activity and durability.
    International Journal of Hydrogen Energy 08/2014; 39(24):12532–12540.

Information

  • Address
    Edgbaston, B15 2TT, Birmingham, United Kingdom
  • Website
    www.birmingham.ac.uk
Information provided on this web page is aggregated encyclopedic and bibliographical information relating to the named institution. Information provided is not approved by the institution itself. The institution’s logo (and/or other graphical identification, such as a coat of arms) is used only to identify the institution in a nominal way. Under certain jurisdictions it may be property of the institution.

3988 Members View all

View all

Top publications last week by downloads

 
Clinical Psychology Review 04/2003; 23(2):225-45.
614 Downloads
 
12/2007: pages 35-50;
202 Downloads

Top Collaborating Institutions

Collaborations

This map visualizes which other institutions researchers from University of Birmingham have collaborated with.

Rg score distribution

See how the RG Scores of researchers from University of Birmingham are distributed.