 |
|
 | Research Interests |  |
|
| The Cross-Entropy Method |
| | The Cross-Entropy (CE) method has been described as ‘one of the most significant developments in randomised optimisation and simulation in recent years.’ In a nutshell, the CE methods involves an iterative procedure where each iteration can be broken down into two phases: (a) generate a randon data sample (trajectories, vectors, etc.) according to a specific mechanism; (b) update the parameters of the randdom mechanism based on this data in order to produce a ‘better’ sample in the next iteration. I am one of the pioneers of the CE method. The simplicity and versatility of the method is explained in my book with R.Y. Rubinstein:
The Cross Entropy Method: A Unified Approach to Combinatorial Optimisation. Monte-Carlo Simulation, and Machine Learning, Springer Verlag, 2004.
The CE method has been applicatied to problems in reliability theory, buffer allocation, telecommunication systems, neural computation, control and navigation, DNA sequence alignment, scheduling and Max-Cut and bipartition problems. |
| |
| Operator Analytic Methods in Telecommunication Systems |
| | Many systems in information technology and telecommunications evolve under conditions of uncertainty. In this context, mathematical modelling is an essential component of the design process. This ARC Discovery Project is focused on providing techniques for analysing operanalytic models, which can be used to study many of the above-mentioned systems. The project aims to deliver new theory and efficient numerical algorithms that will make possible practical analysis of operator-analytic models. |
|
|
