i++ Departmental Newsletter

Week commencing 11 August 2008

Previous Newsletters

GRADUATION 2008 - Official Graduation Pictures

Click on the picture below for more images from this year's Graduation Ceremony, which took place on 18 July in the Ivor Crewe Lecture Hall.

Dr Sam Steel, Head of Department, addresses the graduands at the ceremony

New Prizes Announced

Two new prizes are being introduced for the coming academic year.

·The Active Web Solutions Prize, worth £250, is awarded to the student achieving the highest mark for the final year BSc project. 

·The Telecom Technologies Prize, donated by Telecom Technologies Limited, is awarded by the Board of Examiners to the undergraduate student achieving the highest aggregate mark on the Telecommunication Systems module (EE308-6-AU).  The value of the prize is £150.

Paper Published

Epameinondas Gasparis, Jonathan Nicholson, Amnon H. Eden, Rick Kazman. “Navigating Through the Design of Object-Oriented Programs”. 15th Working Conf. on Reverse Engineering—WCRE (15–18 Oct. 2008), Antwerp, Belgium

Abstract:

The Design Navigator is a tool for reverse-engineering object-oriented programs into charts at any level of abstraction. The Design Navigator discovers the object-oriented building blocks in the design of pro-grams and visualises them in terms of LePUS3, a for-mal Design Description Language. We demonstrate that program visualization in a formal language is not only possible in principle but also of practical benefit.

Paper Announced

Simon M. Lucas, "Reinforcement Learning with Interpolated Table Functions", UKCI 2008 

Abstract:

This paper introduces a new function approximation architecture especially well suited to reinforcement learning problems.   The architecture is based on using sets of interpolated table look-up functions. These offer rapid and stable learning, and are extremely efficient. An empirical investigation is conducted to test their performance on a supervised learning task, and on the mountain car problem, a standard reinforcement learning benchmark. In each case, the interpolated table functions offer extremely competitive performance.

Forthcoming Seminar

Friday 22 August, 2.30-3.00pm, room 4B.531
 
Speaker: Professor Hisao Ishibuchi (Osaka Prefecture University, Japan)

Evolutionary Multiobjective Optimization: Current Research Issues and Machine Learning Applications

Abstract:

Evolutionary multiobjective optimization (EMO) is one of the most active research areas in the field of evolutionary computation. In his talk, he will first demonstrate the popularity of EMO research by showing some statistics such as the number of EMO papers and their citations. Then he will introduce some basic concepts and some common mechanisms used in Pareto-based EMO algorithms. He will explain why Pareto-based EMO algorithms do not work well on many-objective optimization problems. He will introduce some representative approaches and discuss their effectiveness for dealing with many objectives.  After discussions on pure EMO algorithms, he will discuss the hybridisation of EMO algorithms with local search (MOMA). Some research issues on MOMAs will be covered in his talk (e.g., implementation of local search for multiobjective optimization, balance between local search and genetic search, incorporation of problem-specific heuristics, etc.). Finally he will briefly explain how knowledge extraction can be formulated as a multiobjective optimisation problem. He will demonstrate the advantages of the multiobjective approach over the single-objective one by applying them to the design of fuzzy rule-based classification systems.