About the Project
OVERVIEW OF THE RESEARCH PROJECT:
Reinforcement Learning is a rapidly developing field of artificial intelligence (AI). It has already been used successfully in a broad range of problems, including automated game playing, equalling and beating expert players.
This project looks at using reinforcement learning to emulate the way a human being approaches the design of Printed Circuit Boards (PCBs). For over 50 years, automation of design has relied on traditional optimisation approaches, focussed on specific areas. However, there is almost limitless scope to develop more holistic solutions using AI. The objective is to determine how to place components and route circuits across multiple layers to produce designs that are comparable (or superior) in quality to those produced by human experts. This project will explicitly consider the Reinforcement Learning constraints inherent in the domain, including the lack of a well-defined reward function.
It is hoped to develop learning algorithms that are applicable broadly to a wide range of sequential decision problems. In this context, environments with large action spaces, long horizons and delayed rewards are still a considerable challenge for reinforcement learning algorithms. We are particularly keen to investigate hierarchical reinforcement learning approaches. Multi-agent reinforcement learning could be useful as well. There will be substantial flexibility around the approaches that will be developed and tested. These will be influenced by the background and the intellectual interests of the student. Target publication venues are leading conferences and journals in machine learning, including ICML, NeurIPS, and ICLR.
The project will be supervised by Dr Özgür Şimşek, Head of the AI Research Group at the University of Bath’s Department of Computer Science. The university is in the UNESCO World Heritage city of Bath, providing a vibrant research environment in one of the most beautiful areas in the UK. Substantial project support will be available from Zuken Limited, located in nearby Bristol. Zuken has a long track record of technological innovation and is currently looking into machine learning technology to improve its PCB design automation tools for a global customer base. Throughout the project duration, Zuken Limited will provide access to their domain expertise, data sets, and existing software solutions.
Desirable qualities include a strong academic background, intellectual curiosity, and a keen interest in developing machine learning solutions for industrial problems. Extensive programming experience is a plus.
Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree, or the equivalent from an overseas university. A master’s level qualification would be advantageous.
Non-UK applicants must meet our English language entry requirement http://www.bath.ac.uk/study/pg/apply/english-language/index.html.
ENQUIRIES AND APPLICATIONS:
Informal enquiries are welcomed and should be addressed to Dr Özgür Şimşek, email [email protected].
Formal applications should be made via the University of Bath’s online application form for a PhD in Computer Science:
Please ensure that you quote the supervisor’s name and project title in the ‘Your research interests’ section.
More information about applying for a PhD at Bath may be found here:
NOTE: Applications may close earlier than the advertised deadline if a suitable candidate is found; therefore, early application is strongly recommended.
ANTICIPATED START DATE:
Most PhD students start at the beginning of the academic year, on 28 September 2020, but we have flexibility to accommodate other start dates. For more information, please contact the lead supervisor.
Unfortunately, candidates who are classed as 'Overseas' fee-paying purposes will not be considered unless they can provide evidence of their ability to fully self-fund their studies (Overseas tuition fees, bench fees and living costs).
Why not add a message here
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.
Making an e-Learning Tutorial Environment for Mathematics and/or Engineering “Intelligent” :The Design, Implementation and Evaluation of a Tutorial System that Learns from Students’ Mistakes