Currently the ASAP group has THREE fully-funded studentships available: http://www.nottingham.ac.uk/jobs/currentvacancies/ref/SCI1218
The topic (of the 4 listed) for working with me is:
Functional and other Declarative Programming Techniques for Optimisation Problems
Computational search methods for real world problems are often implemented using programming techniques that are too time consuming and costly; especially if parallel/distributed solutions are needed. This project will undertake research to improve the representation and solution of optimisation problems. The focus will be on the use of declarative programming techniques, and, in particular, functional programming (Haskell-family, OCaml, etc.) and constraint programming. The aim will be to represent and solve common problems of practical interest, but in ways that both exploit and enhance techniques in declarative programming.
The intent that it will be an opportunity for someone with a keen-interest in declarative programming techniques (functional, logic, maybe constraints) to investigate ways in which the can be useful tools for real-world problems. Often, in real-world optimisation problems, there are many viable algorithms; however, the sever bottleneck in implementing them is in representing the problems and building solutions - and doing so in a fast, effective, maintainable, reusable fashion (as opposed to a one-off program that is expensive to develop and maintain). It is also of increasing importance for solutions to be make good use of parallel/distributed computation. For these reasons declarative languages should have many advantages, and the project is to explore and demonstrate this. It is more focussed on better programming tools and methods than on the development of new algorithms (though this might well happen as a side-effect.). Of particular interest is the use of monad and similar techniques for the rapid implementation of domain specific languages that could be applied in areas such as healthcare, transport, energy, etc. An example of a class of algorithms to be implemented is dynamic programming as these are heavily used in optimisation and also a good topic for application of FP ideas, such as monads, so as to exploit parallel architectures and parallel coding. The hope is that the requirements of OR would also exploit and improve the design of programming languages.
Although officially based in the ASAP group, the intention is that the project will closely interact with members of the Functional Programming group.
If you have any interest in this, then please feel free to email me: ajp 'at' cs.nott.ac.uk
Note that the application deadline is Feburary 28th, however, the application process is all online, free, and relatively quick and easy, so just apply if interested but not sure. If in doubt, then please feel free to email me.