Visitors are most welcome.
If you are interested in attending or would like to be added to
our mailing list, please contact
Sha Wang.
- Wednesday 19 September at 12:00 in room CS/419: John Woodward
-
Genetic Programming (GP) has been around for 25 years. Genetic Improvement (GI) is new. GI evolves source code, rather than a simulation of code. In other words, GI operates directly on Java or C, for example, whereas GP operates on a tiny instruction set defined by the function set and terminal set. Another fundamental difference is that GI starts with real-world software, whereas GP typically tries to evolve programs from scratch.These differences may not seem important; however this subtle difference opens new possibilities for research. Furthermore we can optimize the physical properties of code such as power consumption, size of code, bandwidth, and other non-functional properties, including execution time.
This tutorial is of interest to people with a GP background interested in applying their techniques to real source code, and software practitioners interested in using automated techniques to improve software. We will not assume prior knowledge of GP.
- Wednesday 10 October at 12:00 in room CS/419: John Woodward
-
How To Write.
When I was a PhD student, I loved research but hated writing it.
Since then I have attended numerous course on writing and come to love it.
I will pass on the tips and tricks I have learnt over the years.
It will also be an opportunity to hear any other views too.
- Wednesday 17 October at 12:00 in room CS/419: Jun Chen
-
An Integrated Routing and Guidance System for Trajectory-based Airport Ground Operations.
The aviation industry has been experiencing a rapid development and the worldwide air traffic is predicted to increase 1.5 times by 2035. This growth is putting the airports, many of which are operating near their maximum capacity, under even more pressure. A more efficient coordination of aircraft movements on the airport surface plays a key role within this context. Recently, new operational concepts and technologies have been proposed to advocate the introduction of the trajectory-based taxi operation (TBTO) as a way of utilising the existing airport infrastructure in a more efficient way. New approaches to integrate the TBTO and guidance systems include guiding pilots via the dynamically lit taxiway lights (also known as "Follow-the-green") and improved flight decks. The number and timing of the lit taxiway lights can convey spatial and temporal information contained in trajectories to pilots. Compared to flight deck automation systems, 'Follow-the-green' has less impact on pilots' workload when it is used to convey the planned trajectories to them. This talk will present some of the preliminary results of integrating "Follow-the-green' and routing and scheduling systems.
- Wednesday 24 October at 12:00 in room CS/419: Michal Weiszer
-
Modelling approaches for Multi-objective Routing and Scheduling for Airport Ground Movement.
Recent research on airport ground movement introduced an Active Routing framework to support multi-objective trajectory-based operations. This results in edges in the airport taxiway graph having multiple costs
such as taxi time, fuel consumption and emissions. In such a graph, multiple edges exist between two nodes
reflecting different trade-offs among the multiple costs. Results suggest that a reduction to a simple graph is sufficient in most cases to provide fast computational time
without compromising the quality of solutions when compared with the results obtained by the
multi-graph approach. The hypothesis is that, the best approach depends on the conditions. When edges
have tight time windows such as during high traffic conditions, higher number of edges could improve
the quality of solutions. The number of edges could be varied for different aircraft or scenarios. In this seminar, I would like to discuss how to test the proposed hypothesis.
- Wednesday 31 October at 12:00 in room CS/419: Sha Wang
-
Sha will present her PhD research plan,
Heuristic search for airport slot allocation problems.
With the expected continued growth in air transportation, the mitigation of the operational efficiency and consequent delays is becoming more and more important. Slot allocation as a means of demand management in multiple congested airports has a significant impact on airport operations at a network level. This requires sophisticated approaches, to intelligently allocate scarce airport resources to unevenly distributed traffic demand for the use of airport facilities. My research aims at developing more intelligent heuristic and hyper-heuristic methods to solve the single airport and network-wide slot allocation problems.
- Thursday 1 November at 16:00, Bancroft Road Teaching Room 3.01:
Ken Reid, Department of Computer Science, University of Stirling.
-
Mixed Metaheuristics for Solving Real World Employee Rostering and Shift Scheduling Problems: A Thesis in Short.
Ken Reid is a final year DAASE PhD student at the University of Stirling, Scotland. Presented in this seminar is a summary of his work over 4 years, which focuses on a variety of solutions to a real world shift scheduling and employee rostering problems. In collaboration with BT, software was also produced, and this will be described. Two papers, one using Variable Neighbourhood Search and another utilising Evolutionary Ruin & Stochastic Recreate, were produced to generate close to optimal shift schedules and employee rosters. Finally Ken will also discuss his next two papers (one still in writing, another currently in consideration) which will summarize his thesis.
- Wednesday 14 November at 13:00 (note the different time): Pol Arias Melia, room to be confirmed.
-
Tutorial: How to implement a mathematical model.
Mathematical models are widely used in the OR literature as they provide a very good understanding of the problem to solve and the possibility to solve it to optimality. In this tutorial, we will learn how to understand the mathematical formulation given in this paper for the Set Packing Problem (SPP), and solve it using Gurobi in Java. We will explain step by step how to transform a series of equations to a runnable program and learn the basics of this process.
- Wednesday 21 November at 12:00 in G. O. Jones building, room UG1: Eun-Seok Kim
-
Scheduling with step-improving processing times.
Scheduling is a decision-making process which deals with the allocation of resources to tasks over given time horizon to optimise one or more objectives, and it has become a major field within operational research with several hundred publications appearing each year. In this talk, we address a single machine scheduling problem with step-improving processing times. For step-improving processing times, job processing times reduce by a job-dependent amount after a common critical date. For the considered problem, we discuss tractability, optimal and heuristic algorithms, and computational experiments of the proposed algorithms. Finally, we discuss some opportunities for future research.
- (Rescheduled) Thursday 22 November at 12:00, Bancroft Building room 3.21: Cunhua Pan
-
By deploying more access points (APs) in fixed area to support the massive machine to machine communications, ultra-dense network (UDN) has been envisioned as one of three vital techniques in the fifth generation (5G) communications.However, severe interference due to the frequency reuse used is the major drawback in UDNs that precludes their application. Thanks to the recent advancement of cloud computing techniques, deploying the UDN under cloud radio access network (CRAN) architecture can effectively address the interference by exploiting the advanced cooperative transmission technique and artificial intelligence (AI) at the central BBU pool. This new network architecture is named as UD-CRAN, which is a brand new paradigm that is in its infancy. Different from conventional CRAN, the number of RRHs is extensive, which yields new challenging technical and practical issues. It is imperative to handle these challenges to make UD-CRANs commercialized, which motivates the application of this project.
- Wednesday 28 November at 12:00 in Bancroft Road Teaching Rooms, room 3.01: Simon Rawles
-
Antiunification for genetic programming.
Over the last few months I've been experimenting with using a suite of
genetic programming operators based around antiunification. The idea is
that similar fragments of code in the program tree will be replaced with
a more general abstraction that concisely expresses them all. By doing
this repeatedly, I expect programs to evolve which take advantage of
higher programming constructs to decompose and explain the problem at
hand. I will present a work in progress and welcome ideas for future
directions.
- Wednesday 5 December at 12:00 in Bancroft Road Teaching Rooms, room 3.01: Yujian Gan
-
Introduction to Machine Learning 1.
In this seminar I attempt to answer the following questions:
(1) What is the difference between neural networks (NN) and genetic
algorithms (GA)?
(2) What is common between NN and GA?
(3) Why are NN and GA machine learning?
(4) What general steps should people follow when applying machine learning?
I will give an example problem that will be solve separately by NN and
GA. Besides this, I will introduce the deep learning framework MXNET.
- Wednesday 12 December at 14:00 in Bancroft Road Teaching Rooms, room 4.02: Prof William Langdon
-
Genetic Improvement by Evolving Program Data.
Genetic Improvement (GI) uses search based software engineering (SBSE)
techniques, such as genetic programming (GP), to evolve existing
programs. It has been used to automatically fix bugs (APR), speed up
code and reduce energy consumption and memory foot print. Usually GI is
applied to optimise source code but recently we have been applying
evolutionary computing (EC) to data within programs.
I will describe 1) updating 50000 integers inside C code to give more
accurate predictions of RNA structure and 2) creating new cube root and
log2 functions with CMA-ES applied to 1024 values in a GNU C library
PowerPC implementation of sqrt().
-
Bill Langdon
is a professorial research associate in CREST at UCL. He
has extensive expertise and experience in genetic programming (GP),
including three books and recently co-chaired the genetic improvement
(GI) workshops --- http://geneticimprovementofsoftware.com/
Bill also has considerable experience of practical software
engineering, having worked for 13 years in industry.
His work has always combined both theory and application. Recent
theoretical analysis has included analysing GP in terms of elementary
landscapes, reversible computing, amorphous computing and The Halting
Problem. Practical work includes evolving multi-classifier systems,
evolving biological motifs to match DNA sequences and predict breast
cancer, exploiting graphics hardware (GPGPU), benchmarks and evolving
improved versions of real-world programs.
- Wednesday 9 January at 14:00 in Bancroft Road Teaching Room 4.01: Yujian Gan
- Introduction to Machine Learning 2: Structural learning.
In this seminar, Yujian will explain what structural learning is,
describe how to solve structural learning in three steps, and give an
example of a structural learning problem.
- Wednesday 16 January at 14:00 in Bancroft Road Teaching Room 4.01: Sachchida Nand Chaurasia
-
A hybrid heuristic for dominating tree problem.
Given an undirected, connected, edge-weighted graph, the dominating tree problem (DTP) seeks on this graph a tree of minimum weight such that each node of the graph either belongs to the tree or is adjacent to a node in the tree. This problem is NP-hard. In this paper, we present an evolutionary algorithm with guided mutation (EA/G) to solve the DTP. This problem has several practical applications in the field of wireless sensor networks. EA/G is a recently proposed evolutionary algorithm that tries to overcome the shortcomings of genetic algorithms (GAs) and estimation of distribution algorithms both and has the characteristics of both.
This work has been published in Springer's Soft Computing Journal.