How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
In order to learn effective behaviour in multi-agent systems, efficient exploration techniques beyond simple, randomised policies are required. My research aims to investigate novel methods with a particular focus on intrinsic rewards as exploration incentives. Such self-assigned rewards serve as additional feedback to motivate guided exploration, which could enable collaborative behaviour for multi-agent reinforcement learning.
October 2019 - present
- Research Assistant
- Designing reinforcement learning (RL) project covering wide range of topics including dynamic programming, single- and multi-agent RL as well as deep RL
September 2017 - October 2017
- - Explained importance of mathematics for CS, formal languages and predicate logic to ~250 participants in daily lectures of the first week - Supervised two groups to provide feedback and further assistance in daily coaching-sessions
October 2016 - March 2017
- Research Assistant
- Taught first-year students fundamental concepts of functional programming, basic complexity theory and inductive correctness proofs in weekly tutorials and office hours
The development of autonomous agents which can interact with other agents to accomplish a given task is a core area of research in artificial intelligence and machine learning. Towards this goal, the Autonomous Agents Research Group develops novel machine learning algorithms for autonomous systems control, with a specific focus on deep reinforcemen...
Successful deployment of multi-agent reinforcement learning often requires agents to adapt their behaviour. In this work, we discuss the problem of teamwork adaptation in which a team of agents needs to adapt their policies to solve novel tasks with limited fine-tuning. Motivated by the intuition that agents need to be able to identify and distingu...
Learning control from pixels is difficult for reinforcement learning (RL) agents because representation learning and policy learning are intertwined. Previous approaches remedy this issue with auxiliary representation learning tasks, but they either do not consider the temporal aspect of the problem or only consider single-step transitions. Instead...
This paper considers how to complement offline reinforcement learning (RL) data with additional data collection for the task of policy evaluation. In policy evaluation, the task is to estimate the expected return of an evaluation policy on an environment of interest. Prior work on offline policy evaluation typically only considers a static dataset....
Deep reinforcement learning (RL) agents that exist in high-dimensional state spaces, such as those composed of images, have interconnected learning burdens. Agents must learn an action-selection policy that completes their given task, which requires them to learn a representation of the state space that discerns between useful and useless informati...
Intrinsic rewards are commonly applied to improve exploration in reinforcement learning. However, these approaches suffer from instability caused by non-stationary reward shaping and strong dependency on hyperparameters. In this work, we propose Decoupled RL (DeRL) which trains separate policies for exploration and exploitation. DeRL can be applied...
Multi-agent deep reinforcement learning (MARL) suffers from a lack of commonly-used evaluation tasks and criteria, making comparisons between approaches difficult. In this work, we evaluate and compare three different classes of MARL algorithms (independent learners, centralised training with decentralised execution, and value decomposition) in a d...
Exploration in multi-agent reinforcement learning is a challenging problem, especially in environments with sparse rewards. We propose a general method for efficient exploration by sharing experience amongst agents. Our proposed algorithm, called Shared Experience Actor-Critic (SEAC), applies experience sharing in an actor-critic framework. We eval...
Multi-agent reinforcement learning has seen considerable achievements on a variety of tasks. However, suboptimal conditions involving sparse feedback and partial ob- servability, as frequently encountered in applications, remain a signiﬁcant challenge. In this thesis, we apply curiosity as exploration bonuses to such multi-agent systems and analyse...
This thesis transfers and evaluates the work of Action Schema Networks (ASNets) for domain-dependent policy learning for classical automated planning. First, we will intro- duce the foundational background of automated planning and deep learning in the form of neural networks. Subsequently, the structure and learning of ASNets will be explained as...
Research intrinsic rewards as exploration incentives to motivate guided, collaborative exploration for multi-agent reinforcement learning.