New Zealand


Training a multi-agent reinforcement learning (MARL) model with a sparse reward is generally difficult because numerous combinations of interactions among agents induce a certain outcome (i.e., success or failure). Earlier studies have tried to resolve this issue by employing an intrinsic reward to induce interactions that are helpful for learning an effective policy. However, this approach requires extensive prior knowledge for designing an intrinsic reward. To train the MARL model effectively without designing the intrinsic reward, we propose a learning-based exploration strategy to generate the initial states of a game. The proposed method adopts a variational graph autoencoder to represent a game state such that (1) the state can be compactly encoded to a latent representation by considering relationships among agents, and (2) the latent representation can be used as an effective input for a coupled surrogate model to predict an exploration score. The proposed method then finds new latent representations that maximize the exploration scores and decodes these representations to generate initial states from which the MARL model starts training in the game and thus experiences novel and rewardable states. We demonstrate that our method improves the training and performance of the MARL model more than the existing exploration methods.

AAMAS 2022

REMAX: Relational Representation for Multi-Agent Exploration

multi-agent exploration

cooperation

multi-agent reinforcement learning

AAMAS (International Conference on Autonomous Agents and Multiagent Systems) is the largest and most influential conference in the area of agents and multiagent systems. The aim of the conference is to bring together researchers and practitioners in all areas of agent technology and to provide a single, high-profile, internationally renowned forum for research in the theory and practice of autonomous agents and multiagent systems. AAMAS is the flagship conference of the non-profit International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS).

The AAMAS conference series was initiated in 2002 in Bologna, Italy as a joint event comprising the 6th International Conference on Autonomous Agents (AA), the 5th International Conference on Multiagent Systems (ICMAS), and the 9th International Workshop on Agent Theories, Architectures, and Languages (ATAL).

Subsequent AAMAS conferences have been held in Melbourne, Australia (July 2003), New York City, NY, USA (July 2004), Utrecht, The Netherlands (July 2005), Hakodate, Japan (May 2006), Honolulu, Hawaii, USA (May 2007), Estoril, Portugal (May 2008), Budapest, Hungary (May 2009), Toronto, Canada (May 2010), Taipei, Taiwan (May 2011), Valencia, Spain (June 2012), Minnesota, USA (May 2013), Paris, France (May 2014), Istanbul, Turkey (May 2015), Singapore (May 2016), São Paulo (2017) and Stockholm, Sweden (2018), Montreal (May 2019), Auckland (May 2020, Virtual), London (May 2021, Virtual).
<br>
<br>



A ticket is required to attend this event, please register using the link below:

https://aamas2022-conference.auckland.ac.nz/attending/registration/

Registration Is Required

AAMAS (International Conference on Autonomous Agents and Multiagent Systems) is the largest and most influential conference in the area of agents and multiagent systems. The aim of the conference is to bring together researchers and practitioners in all areas of agent technology and to provide a single, high-profile, internationally renowned forum for research in the theory and practice of autonomous agents and multiagent systems. AAMAS is the flagship conference of the non-profit International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS).

technical paper


Critical sectors of human society are progressing toward the adoption of powerful artificial intelligence (AI) agents, which are trained individually on behalf of self-interested principals but deployed in a shared environment. Short of direct centralized regulation of AI, which is as difficult an issue as regulation of human actions, one must design institutional mechanisms that indirectly guide agents' behaviors to safeguard and improve social welfare in the shared environment. Our paper focuses on one important class of such mechanisms: the problem of adaptive incentive design, whereby a central planner intervenes on the payoffs of an agent population via incentives in order to optimize a system objective. To tackle this problem in high-dimensional environments whose dynamics may be unknown or too complex to model, we propose a model-free meta-gradient method to learn an adaptive incentive function in the context of multi-agent reinforcement learning. Via the principle of online cross-validation, the incentive designer explicitly accounts for its impact on agents' learning and, through them, the impact on future social welfare. Experiments on didactic benchmark problems show that the proposed method can induce selfish agents to learn near-optimal cooperative behavior and significantly outperform learning-oblivious baselines. When applied to a complex simulated economy, the proposed method finds tax policies that achieve better trade-off between economic productivity and equality than baselines, a result that we interpret via a detailed behavioral analysis.

Adaptive Incentive Design with Multi-Agent Meta-Gradient Reinforcement Learning


Machine learning of Theory of Mind (ToM) is essential to build social agents that co-live with humans and other agents. This capacity, once acquired, will help machine infer the mental states of others from observed contextual action trajectories, enabling future prediction of goals, intention, actions and successor representations. The underlying mechanism for such a prediction remains unclear, however. Inspired by the observation that humans often infer the character traits of others, then use it to explain behaviour, we propose a new neural ToM architecture that learns to generate a latent trait vector of an actor from the past trajectories. This trait vector then multiplicatively modulates the prediction mechanism via the so-called ‘fast weights’ scheme in the prediction neural network, which reads the current context and predicts the future. We empirically show that the fast weights provide a good inductive bias to model the character traits of agents and hence improves mind reading ability. On the indirect assessment of false-belief understanding, the new ToM model enables more efficient helping behaviours.

Learning Theory of Mind via Dynamic Traits Attribution


Reinforcement learning (RL) in partially observable, fully cooperative multi-agent settings (Dec-POMDPs) can in principle be used to address many real-world challenges such as controlling a swarm of rescue robots or a synchronous team of quadcopters. However, Dec-POMDPs are significantly harder to solve than single-agent problems, with the former being NEXP-complete and the latter, MDPs, being just P-complete. Hence, current RL algorithms for Dec-POMDPs suffer from poor sample complexity, thereby reducing their applicability to practical problems where environment interaction is costly. Our key insight is that using just a polynomial number of samples, one can learn a centralized model that generalizes across different policies. We can then optimize the policy within the learned model instead of the true system, reducing the number of environment interactions. We also learn a centralized exploration policy within our model that learns to collect additional data in state-action regions with high model uncertainty. We empirically evaluate the proposed model-based algorithm, MARCO, in three cooperative communication tasks, where it improves sample efficiency by up to 20x. Finally, to investigate the theoretical sample complexity, we propose a tabular model-based method for Dec-POMDPs and show that it achieves sample complexity in polynomial samples.

Centralized Model and Exploration Policy for Multi-Agent RL


We propose a novel formulation of group fairness in the contextual multi-armed bandit (CMAB) setting. In the CMAB setting a sequential decision maker must at each time step choose an arm to pull from a finite set of arms after observing some context for each of the potential arm pulls. In our model arms are partitioned into two or more sensitive groups based on some protected feature (e.g., age, race, or socio-economic status). Despite the fact that there may be differences in expected payout between the groups, we may wish to ensure some form of fairness between picking arms from the various groups. In this work we explore two definitions of fairness: equal group probability, wherein the probability of pulling an arm from any of the protected groups is the same; and proportional parity, wherein the probability of choosing an arm from a particular group is proportional to the size of that group. We provide a novel algorithm that can accommodate these notions of fairness for an arbitrary number of groups, and provide bounds on the regret for our algorithm. We then validate our algorithm using synthetic data as well as two real-world datasets for intervention settings wherein we want to allocate resources fairly across protected groups.

Group fairness in bandit arm selection


Autonomous agents may encounter the transportability problem when they suffer performance deficits from training in an environment that differs in key respects from that in which they are deployed. Although a causal treatment of transportability has been studied in the data sciences, the present work expands its utility into online, multi-agent, reinforcement learning systems in which agents are capable of both experimenting within their own environments and observing the choices of agents in separate, potentially different ones. In order to accelerate learning, agents in these MAT problems face the unique challenge of determining which agents are acting in similar environments, and if so, how to incorporate these observations into their policy. We propose and compare several agent policies that exploit local similarities between environments using causal selection diagrams, demonstrating that optimal policies are learned more quickly than in baseline agents that do not. Simulation results support the efficacy of these new agents in a novel variant of the Multi-Armed Bandit problem with MAT environments.

Exploiting Causal Structure for Transportability in Online, Multi-Agent Environments



In kidney exchange programs, multiple patient-donor pairs each
of whom are otherwise incompatible, exchange their donors to
receive compatible kidneys. The Kidney Exchange problem is
typically modelled as a directed graph where every vertex is either
an altruistic donor or a pair of patient and donor; directed edges are
added from a donor to its compatible patients. The computational
task is to find if there exists a collection of disjoint cycles and paths
starting from altruistic donor vertices of length at most ℓ_𝑐 and ℓ_𝑝
respectively that covers at least some specific number 𝑡 of nonaltruistic vertices (patients). We study parameterized algorithms for
the kidney exchange problem in this paper. Specifically, we design
FPT algorithms parameterized by each of the following parameters:
(1) the number of patients who receive kidney, (2) treewidth of the
input graph + max{ℓ_𝑝, ℓ_𝑐 }, and (3) the number of vertex types in the
input graph when ℓ_𝑝 ≤ ℓ_𝑐 . We also present interesting algorithmic
and hardness results on the kernelization complexity of the problem.
Finally, we present an approximation algorithm for an important
special case of Kidney Exchange.

Parameterized Algorithms for Kidney Exchange

**Read paper on the following link:**

https://ifaamas.org/Proceedings/aamas2022/pdfs/p1563.pdf

**Abstract:**



Optimal Local Bayesian Differential Privacy over Markov Chains


In MARL, communication among agents is essential to establish cooperation. Over the realistic wireless network, many factors can affect transmission reliability, especially considering that the wireless network condition varies with agents’ mobility. We propose a framework that improves the intelligence of communication over realistic wireless networks in two fundamental aspects: (1) When: Agents learn the timing of communication based on message importance and wireless channel condition. We further propose a communication lagging technique to make the training end-to-end differentiable. (2) What: Agents augment message contents with
wireless network measurements. The messages improve both the game and communication actions of the agents. Experiments on a standard environment show that compared with state-of-the-art, our framework enables more intelligent collaboration and thus achieves significantly better game performance, convergence speed and communication efficiency.

Intelligent Communication over Realistic Wireless Networks in Multi-Agent Cooperative Games

Federated learning (FL) allows distributed clients to collaboratively train a global model without sharing their private data. We propose a novel FL algorithm, which can accurately evaluate the contributions of clients and aggregate global model in a more efficient manner. More specifically, we propose to use a Deep Reinforcement Learning (DRL) method to dynamically learn the contributions of clients in each communication round. Based on this, we adaptively assign appropriate weights to clients, which will be used in global model aggregation. By improving the process of global mode aggregation, our proposed scheme greatly improves the performance of federated learning.

Adaptive Aggregation Weight Assignment for Federated Learning: A Deep Reinforcement Learning Approach


PointGoal navigation has seen significant recent interest and progress,
spurred on by the Habitat platform and associated challenge.
In this paper, we study PointGoal navigation under both a sample
budget (75 million frames) and a compute budget (1 GPU for 1 day).
We conduct an extensive set of experiments, cumulatively totaling
over 50,000 GPU-hours, that let us identify and discuss a number of
ostensibly minor but significant design choices – the advantage esti-
mation procedure (a key component in training), and visual encoder
architecture. Overall, these design choices lead to considerable and
consistent improvements. Under a sample budget, performance for
RGB-D agents improves 3 SPL on Gibson (4% relative improvement)
and 20 SPL on Matterport3D (43% relative improvement). Under a
compute budget, performance for RGB-D agents improves by 3 SPL
on Gibson (5% relative improvement) and 15 SPL on Matterport3D
(50% relative improvement). Our findings and recommendations
will serve to make the community’s experiments more efficient – to
reach 50 SPL with RGB-D on Matterport3D, they reduce the samples
needed by 3x and the training time 2x.

REMAX: Relational Representation for Multi-Agent Exploration

Downloads

Next from AAMAS 2022

Adaptive Incentive Design with Multi-Agent Meta-Gradient Reinforcement Learning

Similar lecture

Learning to Transfer Role Assignment Across Team Sizes

REMAX: Relational Representation for Multi-Agent Exploration

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAMAS 2022

Adaptive Incentive Design with Multi-Agent Meta-Gradient Reinforcement Learning

Similar lecture

Learning to Transfer Role Assignment Across Team Sizes

Downloads