Research

Area of Interests and Research Topics

I have great interest in some of these research topics:

  • Multiple stochastic Gaussian processes
  • Compound state-space time-series regression using Gaussian process models
  • Fast algorithms implementation for large-scale Gaussian regressions
  • Identification of nonlinear dynamic systems
  • Analysis of wind turbine aerodynamics and electro-mechanical dynamics to aid and improve design controller
  • Data mining, statistical analysis and specialised reporting
  • Effective data cleansing algorithm using Bayesian technique

My past research was supported by various funding sources, such as the Engineering and Physical Sciences Research Council (EPSRC) and the Science Foundation of Ireland (SFI). The thesis for the research was completed in 2008.

About Hamilton Institute Systems Modelling Group

The Systems Modelling Group (SMG) was established in 2006, led by Professor Bill Leithead, where the focus was on data modelling of nonlinear dynamic systems. Specialised in mathematical research within the context of engineering applications, the group utilised various methodologies, such as Machine Learning, particularly in the use of Gaussian process prior models, for System Identification. The following members were part of the SMG:

  • Professor Bill Leithead (Director of Hamilton Institute, NUIM)
  • Professor Douglas Leith (Director of Industrial Control Centre, University of Strathclyde)
  • Professor Zhang Yunong (Professor, Sun Yat-sen University)
  • Dr. Emanuele Ragnoli (IBM)
  • Dr. Keith Neo (SAP)

Several works were published over the years, and can be found at the Hamilton Institute publication library.

Research Hardware and Equipment

The following are some of the equipments used during my research in Gaussian processes; particularly the development of fast algorithms, which require the use of fast computing machines to perform burn-in training and predictions.

  • Dell Laptop 3.06GHz
  • Dell Desktop 3.06GHz (University of Strathclyde)
  • Dell Desktop 2.8GHz (Hamilton Institute)

These systems were installed with Microsoft Windows XP and Linux (SuSE 9.3 and Ubuntu 6.10) operating systems. Compiled C codes are the preferred choice of programming languages for both platforms.

Leave a Reply

Your email address will not be published. Required fields are marked *

*