Connectionists: PhD Studentship in Robot Multi-Objective Reinforcement Learning

[Matthias Rolf]

Qualification Type: 	PhD
Location: 		Oxford
Funding for: 		UK Students, EU Students
Funding amount: 	£16,536 per year, which will increase by inflation each 
year.
Hours: 			Full Time

The School of Engineering, Computing and Mathematics at Oxford Brookes 
University is seeking an independent and highly motivated candidate from 
a science or technology discipline. The candidate should have an 
interest in robot learning and have knowledge of machine learning methods.

Brief Project Description

This project will develop new methods of multi-objective reinforcement 
learning specifically tailored to robotic problems such as navigation or 
manipulation. Current methods of multi-objective and risk averse 
reinforcement learning will be combined and developed further to give a 
reinforcement learning robot the capability of striking the right 
balance between its objectives, and ensuring safe interaction with the 
physical and social world next to its primary task objective. The 
project’s main goal is to develop novel algorithms and evaluate them in 
the context of existing physical robotic platforms currently available 
to the AI and Robotics group in the school.

The essential selection criteria include:

- At least an upper second class degree (preferably MSc) in a Science or 
Technology discipline
- Knowledge of, and experience in, machine learning
- Keen interest in robotics
- Strong mathematical abilities
- Capability to work independently and as part of a team
- Excellent written and oral communication and organisational skills. 
Proficiency in written English is required.
- A real passion and commitment for research.

The successful candidate will:

- Develop novel multi-objective reinforcement learning methods tailored 
to real-world robotic learning problems
- Formulate and implement robotic test scenarios for reinforcement 
learning on existing robotic platforms in the school
- Experimentally assess and analyze the performance of the learning 
algorithms on physical robotic platforms in relation to the state of the art
- Participate in research meetings, travel to partner institutions and 
undertake supplementary research methods training.

Application Process:

Please request an application pack from the Research Support Office, 
tdestudentships at brookes.ac.uk, quoting “PhD Studentship in Robot 
Multi-Objective Reinforcement Learning”.

Fully completed applications must be sent to 
tdestudentships at brookes.ac.uk, by *17.00 on Friday 7th February 2020*. 
As part of the application process you will be required to upload a 
cover letter, a brief statement of research interests (describing how 
past experience and future plans fit with the advertised position), CV 
and the details of two referees as part of your online application. 
Please be advised that the selection process will involve an interview 
during February.

Information:

For all informal requests about the project and this studentship, please 
contact Dr Matthias Rolf (mrolf at brookes.ac.uk).