Back to search results

PhD Studentship: Collective Intelligence in Multi-Agent Reinforcement Learning for Deliberative Processes

University of Birmingham - School of Computer Science

Qualification Type: PhD
Location: Birmingham
Funding for: UK Students, EU Students
Funding amount: £17,668 pa and tuition fees of £4,620 pa
Hours: Full Time
Placed On: 31st January 2023
Closes: 28th February 2023

Reinforcement Learning (RL) has achieved exceptional success in recent years, especially for sequential decision-making and tasks that require continuous control. Examples include the game of Go, video games – especially strategy games such as StarCraft – and also robotics.

Recently, a prominent area of research involves the extension to multi-agent reinforcement learning (MARL). In this project, we focus on decentralised MARL, i.e., where agents do not intrinsically know the state of the other agents but can interact with one another. The main advantages of this approach include faster learning (e.g., through parallel computation), robustness to individual failures, transfer learning from more experienced agents, increasing number of applications (both in cooperative and adversarial settings), to name a few.

However, the extension of RL to multi-agent settings brings several challenges. One of the main challenges is the coordination of these agents. Another fundamental issue is scalability and, therefore, the communication between large numbers of agents. Recent literature has tackled the scalability issue through the formulation of optimality guarantees in the limit of an infinite number of agents through a mean-field approach [J01]. The advantage of this approach is to be able to quantify upper bounds in the performance of a large number of agents when performing collective decision-making [J01]. 

However, when translating these results back to the original problem with finite population, the interactions among the agents play a crucial role and similar approaches do not scale down well. Therefore, working with multi-agent systems is necessary to ensure optimality.

The aim of this project is to study multi-agent reinforcement learning by embedding elements of game theory to tackle collective decision-making. Specifically, three objectives are considered:

  1. To design and develop scalable and robust MARL models for collective decision-making in complex tasks. This requires the investigation of diversified approaches to learning, including formalism from bio-inspired game theoretic approaches [J02, J03].
  2. To investigate a range of communication policies to foster coordination between the agents. This involves the study of the impact of heterogeneity in the agents’ interactions for scalability and robustness. These interactions can be captured by regular networks or scale-free networks and can drastically change the behaviour of the system and its properties [J02, J04].
  3. To analyse the developed framework through a mean-field game approach [J01]. This leads to a much more tractable analysis for robustness in cooperative and competitive contexts [J01].

Eligibility:

  • Strong mathematical background.
  • Strong background in machine learning and multi-agent systems.
  • Good understanding of game theoretic concepts.
  • Strong programming skills in Python (and MATLAB), experience using ML toolboxes (Pytorch/Google Colab/PettingZoo).
  • Excellent communication skills.

Eligibility: First or Upper Second Class Honours undergraduate degree and/or postgraduate degree with Distinction (or an international equivalent). We also consider applicants from diverse backgrounds that have provided them with equally rich relevant experience and knowledge. Full-time and part-time study modes are available. UoB studentships are open to all and we particularly welcome applications from under-represented groups, including, but not limited to BAME, disabled and neuro-diverse candidates. We also welcome applications for part-time study.

Funding

The position is offered is for three and a half years full-time study. The value of the award is a stipend of £17,668 (subject to review) pa and tuition fees of £4,620 pa. Awards are usually incremented on 1 October each following year.

We value your feedback on the quality of our adverts. If you have a comment to make about the overall quality of this advert, or its categorisation then please send us your feedback
Advert information

Type / Role:

Subject Area(s):

Location(s):

PhD tools
 

PhD Alert Created

Job Alert Created

Your PhD alert has been successfully created for this search.

Your job alert has been successfully created for this search.

Ok Ok

PhD Alert Created

Job Alert Created

Your PhD alert has been successfully created for this search.

Your job alert has been successfully created for this search.

Manage your job alerts Manage your job alerts

Account Verification Missing

In order to create multiple job alerts, you must first verify your email address to complete your account creation

Request verification email Request verification email

jobs.ac.uk Account Required

In order to create multiple alerts, you must create a jobs.ac.uk jobseeker account

Create Account Create Account

Alert Creation Failed

Unfortunately, your account is currently blocked. Please login to unblock your account.

Email Address Blocked

We received a delivery failure message when attempting to send you an email and therefore your email address has been blocked. You will not receive job alerts until your email address is unblocked. To do so, please choose from one of the two options below.

Max Alerts Reached

A maximum of 5 Job Alerts can be created against your account. Please remove an existing alert in order to create this new Job Alert

Manage your job alerts Manage your job alerts

Creation Failed

Unfortunately, your alert was not created at this time. Please try again.

Ok Ok

Create PhD Alert

Create Job Alert

When you create this PhD alert we will email you a selection of PhDs matching your criteria.When you create this job alert we will email you a selection of jobs matching your criteria. Our Terms and Conditions and Privacy Policy apply to this service. Any personal data you provide in setting up this alert is processed in accordance with our Privacy Notice

Create PhD Alert

Create Job Alert

When you create this PhD alert we will email you a selection of PhDs matching your criteria.When you create this job alert we will email you a selection of jobs matching your criteria. Our Terms and Conditions and Privacy Policy apply to this service. Any personal data you provide in setting up this alert is processed in accordance with our Privacy Notice

 
 
 
More PhDs from University of Birmingham

Show all PhDs for this organisation …

More PhDs like this
Join in and follow us

Browser Upgrade Recommended

jobs.ac.uk has been optimised for the latest browsers.

For the best user experience, we recommend viewing jobs.ac.uk on one of the following:

Google Chrome Firefox Microsoft Edge