A leading academic and a recent graduate from the University's Department of Computing Science have won an award at a major national conference for specialists in artificial intelligence (AI) and machine learning research.
Senior lecturer Wei Pang and Edvinas Byla, who graduated with a First-class Honours in Computing Science this year, won the best paper award at the 19th Annual UK Workshop on Computational Intelligence (UKCI 2019).
The paper they worked on together – ‘DeepSwarm: Optimising Convolutional Neural Networks using Swarm Intelligence’ – has attracted widespread attention in the machine learning community. The work, which proposes a new system to automate the design of problem-solving AI tools, was based on Edvinas’ honours project, which was supervised by Dr Pang.
Explaining the background to their research, Dr Pang said: “The design of a good AI tool heavily relies on human experience and sometimes their intuitions, and in many cases, it is very time-consuming and requires huge human efforts.
“DeepSwarm is trying to automate this process - given a new problem, DeepSwarm will automatically design deep neural networks which are fit for the problem.
“This will streamline the process and save us a lot of time, and DeepSwarm’s ultimate goal is to help those people who don’t know a lot about AI produce AI systems for solving their own problems.”
Commenting on the award, he added: “UKCI has been a major national conference in the field of computational intelligence since 2001, and this year there were a total of 45 papers accepted.
“Winning the best paper award, particularly when so many high-quality papers were accepted, is testament to the excellent AI research taking place within the Department of Computing Science and showcases the strength of our undergraduate teaching.
“I am proud to have worked alongside Edvinas on this paper, and delighted that our work has been recognised in this way.”