Computing Science and Mathematics
Seminars, 2021/2022
Seminars will take place via Microsoft Teams, with a meeting link to be shared via the seminar-announce emails. Unless otherwise stated, from 15.00 to 16.00 on
Friday afternoons during semester time, followed by informal discussions.
Spring 2022
| Date | Speaker | Title/Abstract |
|---|---|---|
| Friday 28 January |
||
| Friday 4 February |
||
| Friday 11 February |
||
| Friday 18 February |
||
| Friday 25 February |
Dr Ken Reid, Department of Animal Science, Michigan State University | Genomic Prediction using Evolutionary Computation, Machine Learning and Traditional Statistical Approaches The completion of the human genome project was a major milestone for humanity. However, the focus then quickly shifted to the more complex problem of understanding the new data we had access to. Traditionally, selection of animals for breeding specific traits (such as disease resistance or milk yield) was through careful observation of phenotypes. While this traditional approach has been successful, discerning which parts of DNA are causative for specific phenotypic traits has proven to be more effective. While some traits can be easily mapped to specific DNA segments, other traits have proven impossible to map thus far. In this seminar, we will explore these problems and consider various possible solutions for mapping causative quantitative trait loci to phenotypes in various species, while making use of modern hardware and methodologies. |
| Friday 4 March |
No seminar - reading week |
|
| Friday 11 March |
||
| Wednesday 16 March (3-4pm) |
Dr Sandy Brownlee, CSM, University of Stirling | Modelling and Optimisation of Aircraft Ground Movement for Greener, Faster Airport Operations Aircraft are designed to operate efficiently in the air, but a necessary part of any flight is the taxiing of aircraft between terminal and runway at either end of the journey. The allocation of routes to aircraft for their travel along taxiways is known as the Ground Movement problem, and it has a strong influence on the capacity of an airport and environmental impact, with taxiing accounting for a disproportionate level of emissions. The problem is particularly challenging. It has multiple conflicting objectives to optimise (such as minimising taxi time and fuel consumption). It also has considerable uncertainty, which arises from the complex operations of an airport. In this talk I will cover research carried out under the EPSRC funded "TRANSIT" project. I will cover the application of machine learning to accurately predict taxi times for aircraft, including identification of important features that drive the model, and a hybrid optimisation approach, combining evolutionary algorithms and exact short path algorithms, to generate efficient routes. Note - this talk is being hosted by Michigan State University, so will be on a zoom link, to be circulated by email |
| Friday 25 March |
||
| Friday 1 April |
Dr David Li Department of Biomedical Engineering, University of Strathclyde | Advanced time-resolved imaging and spectroscopy techniques for life science Super-resolution time-resolved imaging and spectroscopy techniques can reveal biological processes at the molecular level. They can unravel mechanisms underlying diseases and facilitate drug development. However, we also enter the low light regime (only a few photons are acquired), and new acquisition and analysis approaches are sought after. Speaker Bio: David Li is a senior lecturer in optoelectronics at the Department of Biomedical Engineering, University of Strathclyde. He received his PhD in electrical engineering from National Taiwan University. He then joined the Industrial Technology Research Institute (ITRI), Taiwan, as an R&D engineer working on 1.25-12.5Gbps optical communication chipsets and wireless communication IP, knowledge transfer, and international joint projects. From 2007 to 2011, he worked at the Institute for Integrated Micro and Nano Systems, University of Edinburgh, on the European projects "MEGAFRAME" and "METOXIA" to develop CMOS time-resolved imaging systems for biomedical applications. From January 2014, he joined Strathclyde Institute of Pharmacy & Biological Science, after his first lecturing position in biomedical engineering and embedded systems at the School of Engineering and Informatics, the University of Sussex. He is currently leading a BBSRC project to establish a SuPer-Resolution (< 50nm) time-resolved (< 5ps) ImagiNg and specTroscopy (SPRINT) system and is looking for new applications and collaborators to use SPRINT. |
| Friday 8 April |
||
| Friday 15 April |
||
| Friday 22 April |
||
| Friday 29 April |
||
| Wednesday 11 May (3-4pm) |
Clifford Bohm, Post-Doctoral Student at Michigan State University | Understanding the Evolution of Cognition with Structural and Information-Theory Based Analysis In this talk, I will present work on developing methods to better understand the structure of evolved digital cognitive architectures such as Markov Brain, RNN, and CGP. First, I will present Fragmentation, an information analysis that can be used to identify how data flows through cognitive architectures. Then I will discuss the Comparative Hybrid Approach, an empirical technique that can shed light on how the components of a given architecture may account for differences in performance between architectures. |
| Wednesday 18 May (3-4pm) |
Prof. Gabriela Ochoa | Neuroevolution Trajectories and Landscapes Neuroevolution, the use of evolutionary algorithms to design neural networks, has a long tradition in evolutionary computation with roots in the late 1980s and early 1990s. Most neuroevolution systems optimise both the neural network topology and its weights. However, when scaling up to contemporary deep models with millions of weights for supervised learning tasks, gradient-based weight optimisation outperforms evolutionary methods. In consequence, many recent neuroevolution systems use gradient-based weight optimisation and only evolve the topology. This talk overviews our recent work on modelling neuroevolution systems with search trajectory networks (STNs) and local optima networks (LONs) with the aim of providing a visual and quantitative understanding of search and optimisation in this domain. |
Previous Seminar Series
Top image: Image and vision processing.
Courtesy of Dr. Deepayan Bhowmik.
Computing Science and Mathematics
Faculty of Natural Sciences
University of Stirling
Stirling
FK9 4LA
Scotland
UK