Computing Science and Maths Seminars, 2019/2020
Seminars will take place in Room 4B96, Cottrell Building, University of Stirling. Unless otherwise stated, from 15.00 to 16.00 on Friday afternoons during semester time, followed by discussions and snacks in 4B94. For instructions on how to get to the University, please look here.
Spring 2020
| Date | Speaker | Title/Abstract |
|---|---|---|
| Friday 17th January |
Prof Carron Shankland | Writing workshop for Lovelace 2020 submissions The BCSWomen Lovelace Colloquium is a free, one day conference for women undergraduates and taught masters students and is being held in Stirling in April. All women students of computing and related subjects are invited to enter one of the poster contests. Abstracts are due by 31st January 2020. Need help on how to write an abstract? Come to our workshop! There will be a second writing workshop on Wednesday 22nd January 11-12 in 4B96. |
| Friday 24th January |
Dr Marc Roper, Computer and Information Sciences, University of Strathclyde | Using Machine Learning to Classify Test Outcomes In recent years, software testing research has produced notable advances in the area of automated test data generation. It is now possible to take an arbitrary system and automatically generate volumes of high-quality test data. But the problem of checking the correctness or otherwise of the outputs (termed the "oracle problem") still remains. This talk examines how machine learning techniques can be used to separate failing and passing cases. The feasibility of the approach is demonstrated and shown to have the potential to reduce by an order of magnitude the numbers of outputs that need to be examined following a test run. |
| Friday 31st January |
Dr Joao Mota, Signal and Image Processing Laboratory, Heriot-Watt University | Multi-Modal Data Processing: An Approach via Sparsity Making sense of modern datasets, in which data is often multi-modal and heterogeneous, is a challenging task that is becoming increasingly important for both academia and industry. In this work, we look at sparsity-based approaches to process multi-modal and heterogeneous data. We start with the problem of integrating prior knowledge into sparse reconstruction schemes. Prior information here means a signal similar to the signal to be reconstructed, for example, in medical imaging, a prior scan of the same patient. Our theory provides a minimal number of measurements required to reconstruct the original signal as a function of the quality of the prior information. We then describe an approach to separate the x-rays of the paintings in the door panels of the Ghent Altarpiece, a 15th century art work by Van Eyck which was recently restored. Our method uses the visual images to aid the separation process and outperforms prior state-of-the-art methods, such as morphological component analysis. Finally, we describe a single-image super-resolution method that combines deep convolutional networks with a sparsity-based approach and explore how such a framework can be extended to processing arbitrary, but correlated, heterogeneous data such as images and sound, or text and location. Speaker Bio: Joao F. C. Mota received the M.Sc. degree and the Ph.D. degree in Electrical and Computer Engineering from the Technical University of Lisbon, Portugal, in 2008 and 2013, respectively. He also received the Ph.D. degree in Electrical and Computer Engineering from Carnegie Mellon University, US, in 2013. From 2013 to 2016, he was Senior Research Associate at University College London, London, UK. In 2017, he became Assistant Professor in the School of Engineering and Physical Sciences at Heriot-Watt University, Edinburgh, UK., where he is also affiliated with the Institute of Sensors, Signals, and Systems. His current research interests include theoretical and practical aspects of high-dimensional data processing, multi-modal processing, inverse problems, optimization theory, machine learning, data science, and distributed information processing and control. He was the recipient of the 2015 IEEE Signal Processing Society Young Author Best Paper Award for the paper "Distributed Basis Pursuit", published in IEEE Transactions on Signal Processing. |
| Friday 7th February |
Available Slot | |
| Friday 14th February |
Dr Lilia Georgieva, Department of Computer Science, Heriot-Watt University | Fake or not: the art and science of disambiguation of personal names. Personal name disambiguation or named entity linking (NEL) is the task of linking data for a personal name online to a unique comparable entry in the real world. Algorithms for NEL consist of three main components: extractor, searcher, and disambiguator. Existing approaches for NEL use exact-matched look-up to generate a set of candidate entities for each of the mentioned names. The exact-matched look-up is inadequate due to, for example, a lack of a uniform name representation or language-specific issues. In addition, the performance of disambiguator in ranking candidate entities is limited by context similarity, which is an inflexible and highly variable feature. We propose a new approach that can be used to both identify and disambiguate personal names. Our NEL algorithm uses: as an extractor: a control flow graph, as a searcher: standardized similarity measures, and as a disambiguator: grammar-based entity coherence method. Experimental results, evaluated on real-world data sets, show that accuracy of our NEL is 92%, which is higher than the accuracy of previously used methods. Speaker Bio: Lilia Georgieva is a lecturer in Computer Science at Heriot-Watt University. She has an Executive MBA in Entrepreneurship and Business Engineering (Babson College, 2015) and PhD in Computer Science (University of Manchester, 2003). She co-founded the UK chapter of ACM-Women in Computing and serves as the treasurer of the UK chapter. Currently, she is a theory them leader at the Scottish Informatics and Computer Science Alliance (SICSA) and the vice chair of Future and Emerging Technologies Panel, an advisory group for EU’s Horizon 2020. Her research is in the area of applications of formal methods in knowledge representation and reasoning, natural language processing, program analysis, and security. |
| Friday 21st February |
None | Reading week |
| Friday 28th February |
Internal Event |
|
| Friday 6th March |
Dr Sandy Brownlee, CSM, University of Stirling | Search-based approaches to improving Java programs Writing code is hard: writing good code is considerably harder. Many aspects of code improvement can be formulated as search problems, where we are searching over alternative implementations of particular functionality. We can apply local search and evolutionary algorithms to these problems, getting the computer to do the hard work of improving existing code. An emerging field termed genetic improvement of software has shown that this is surprisingly effective. In this talk I will introduce a toolkit, called Gin, for experimentation around genetic improvement of Java code. I will share some results showing that source code is more robust than might be expected. I will then look at the topic of reducing computational energy consumption, which is particularly important at the extremes (i.e., mobile devices and datacentres). In many cases there is a trade-off between functionality and energy consumption, and I will summarise a few experiments exploring this problem. |
| Friday 13th March |
No seminar this week! |
|
| Friday 20th March |
Postponed due to virus outbreak |
|
| Friday 27th March |
External speaker postponed due to virus outbreak |
|
| Friday 3rd April |
External speaker postponed due to virus outbreak |
|
| Friday 10th April |
Easter - University Closed |
|
| Friday 17th April |
Dr Paul McMenemy, CSM, Stirling | TBA |
| Friday 24th April |
Available slot | |
| Friday 1st May |
Dr Simona Hapca, CSM, Stirling | TBA |
Previous Seminar Series
1996:
Autumn
Top image: Evolution of sounds by crowd-sourcing, from Brownlee, A. E. I., Kim, S-J., Wan, S-H., Chan, S. & Lawson, J. A. Crowd-Sourcing the Sounds of Places with a Web-Based Evolutionary Algorithm. Companion Proc. of the Genetic and Evolutionary Computation COnference 2019, Prague, Czech Republic, pp 131-132. DOI:10.1145/3319619.3322028