Filters








3,985 Hits in 6.0 sec

Multi-Agent Game Abstraction via Graph Attention Neural Network [article]

Yong Liu, Weixun Wang, Yujing Hu, Jianye Hao, Xingguo Chen, Yang Gao
2019 arXiv   pre-print
We integrate this detection mechanism into graph neural network-based multi-agent reinforcement learning for conducting game abstraction and propose two novel learning algorithms GA-Comm and GA-AC.  ...  In this paper, we model the relationship between agents by a complete graph and propose a novel game abstraction mechanism based on two-stage attention network (G2ANet), which can indicate whether there  ...  Figure 1 : 1 Game Abstraction based on two-stage attention mechanism and Graph Neural Network (GNN). Figure 2 : 2 Two-Stage Attention Neural Network.  ... 
arXiv:1911.10715v1 fatcat:mdel7ormavhm5ejintlj5ykpby

Multi-Agent Game Abstraction via Graph Attention Neural Network

Yong Liu, Weixun Wang, Yujing Hu, Jianye Hao, Xingguo Chen, Yang Gao
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We integrate this detection mechanism into graph neural network-based multi-agent reinforcement learning for conducting game abstraction and propose two novel learning algorithms GA-Comm and GA-AC.  ...  In this paper, we model the relationship between agents by a complete graph and propose a novel game abstraction mechanism based on two-stage attention network (G2ANet), which can indicate whether there  ...  Figure 1 : 1 Game Abstraction based on two-stage attention mechanism and Graph Neural Network (GNN). Figure 2 : 2 Two-Stage Attention Neural Network.  ... 
doi:10.1609/aaai.v34i05.6211 fatcat:t5w4zz7o3ndj7hloogfngoc3oq

Learning Multi-agent Action Coordination via Electing First-move Agent [article]

Jingqing Ruan and Linghui Meng and Xuantang Xiong and Dengpeng Xing and Bo Xu
2022 arXiv   pre-print
We propose a novel election mechanism in which we adopt a graph convolutional network to model the interaction among agents and elect a first-move agent for asynchronous guidance.  ...  Learning to coordinate actions among agents is essential in complicated multi-agent systems.  ...  We use multi-head dot-product attention as the convolutional kernel to learn how to abstract the relationship between agents, as described in (Jiang et al. 2020) .  ... 
arXiv:2110.08126v3 fatcat:vckcw3xvmrflbafnc4d34hmfjq

Deep Implicit Coordination Graphs for Multi-agent Reinforcement Learning [article]

Sheng Li, Jayesh K. Gupta, Peter Morales, Ross Allen, Mykel J. Kochenderfer
2021 arXiv   pre-print
DICG consists of a module for inferring the dynamic coordination graph structure which is then used by a graph neural network based module to learn to implicitly reason about the joint actions or values  ...  Multi-agent reinforcement learning (MARL) requires coordination to efficiently solve certain tasks.  ...  ACKNOWLEDGMENTS A.2 StarCraft II Multi-agent Challenge (SMAC) We use the open source SMAC implementation by Samvelyan et al. [32] . We use the SMAC implementation 4 by Samvelyan et al. [32] .  ... 
arXiv:2006.11438v2 fatcat:7kc4f4xabzc5tg4655prfo326y

Relational Deep Reinforcement Learning [article]

Vinicius Zambaldi, David Raposo, Adam Santoro, Victor Bapst, Yujia Li, Igor Babuschkin, Karl Tuyls, David Reichert, Timothy Lillicrap, Edward Lockhart, Murray Shanahan, Victoria Langston, Razvan Pascanu (+3 others)
2018 arXiv   pre-print
In the StarCraft II Learning Environment, our agent achieves state-of-the-art performance on six mini-games -- surpassing human grandmaster performance on four.  ...  It uses self-attention to iteratively reason about the relations between entities in a scene and to guide a model-free policy.  ...  This mechanism has parallels with graph neural networks and, more generally, message passing computations [27, 28, 29, 12, 30] .  ... 
arXiv:1806.01830v2 fatcat:dgkztdki25dhha5i2jgve5n73i

Heterogeneous Graph Attention Networks for Learning Diverse Communication [article]

Esmaeil Seraj, Zheyuan Wang, Rohan Paleja, Matthew Sklar, Anirudh Patel, Matthew Gombolay
2021 arXiv   pre-print
We propose heterogeneous graph attention networks, called HetNet, to learn efficient and diverse communication models for coordinating heterogeneous agents towards accomplishing tasks that are of collaborative  ...  Multi-agent teaming achieves better performance when there is communication among participating agents allowing them to coordinate their actions for maximizing shared utility.  ...  MARL with Graph Neural Networks -Graph Neural Networks (GNNs) are a class of deep neural networks that learn from unstructured data by representing objects as nodes and relations as edges and aggregating  ... 
arXiv:2108.09568v2 fatcat:kqnvqafqmzg3jc3jsbbcuwq5wm

RACA: Relation-Aware Credit Assignment for Ad-Hoc Cooperation in Multi-Agent Deep Reinforcement Learning [article]

Hao Chen, Guangkai Yang, Junge Zhang, Qiyue Yin, Kaiqi Huang
2022 arXiv   pre-print
RACA takes advantage of a graph-based relation encoder to encode the topological structure between agents.  ...  In recent years, reinforcement learning has faced several challenges in the multi-agent domain, such as the credit assignment issue.  ...  As a kind of graph neural network, graph convolutional neural network (GCN) is often used to process structured input data and incorporate neighborhood information.  ... 
arXiv:2206.01207v1 fatcat:odo2iudgnzfatgbd5d2d2gi4ui

Deep Interactive Motion Prediction and Planning: Playing Games with Motion Prediction Models [article]

Jose L. Vazquez, Alexander Liniger, Wilko Schwarting, Daniela Rus, Luc Van Gool
2022 arXiv   pre-print
This work presents a module that tightly couples these layers via a game-theoretic Model Predictive Controller (MPC) that uses a novel interactive multi-agent neural network policy as part of its predictive  ...  Finally, we show that our multi-agent policy network learns to drive while interacting with the environment, and, when combined with the game-theoretic MPC planner, can successfully generate interactive  ...  Modeling interactions between all the agents in the scene has been addressed by using multi-headed attention models (Mercat et al., 2020; Rella et al., 2021) or by using Graph Neural Networks (GNN)  ... 
arXiv:2204.02392v1 fatcat:gxmb5wb4kngibdco4o4c5iiley

MAPEL: Multi-Agent Pursuer-Evader Learning using Situation Report

Sagar Verma, Richa Verma, P.B. Sujit
2019 2019 International Joint Conference on Neural Networks (IJCNN)  
We use Recurrent Neural Networks (RNNs) to parameterize the spatio-temporal graph.  ...  We provide empirical results to show how agents cooperate under these two methods. Index Terms-multi-agent learning; deep reinforcement learning; recurrent neural network I.  ... 
doi:10.1109/ijcnn.2019.8852457 dblp:conf/ijcnn/VermaVS19 fatcat:k5jghmyv5zgglh227uvuo5osu4

MAPEL: Multi-Agent Pursuer-Evader Learning using Situation Report [article]

Sagar Verma and Richa Verma and P.B. Sujit
2019 arXiv   pre-print
We use Recurrent Neural Networks (RNNs) to parameterize the spatio-temporal graph.  ...  We present Multi-Agent Pursuer-Evader Learning (MAPEL), a class of algorithms that use spatio-temporal graph representation to learn structured cooperation.  ...  We use Recurrent Neural Networks (RNNs) to parameterize the spatio-temporal graph.  ... 
arXiv:1910.07780v1 fatcat:gl6fji3fbzcalpdmqtnihmvbnm

Enhanced Random Forest Algorithms for Partially Monotone Ordinal Classification

Christopher Bartley, Wei Liu, Mark Reynolds
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Ideally a solution would (a) maximise accuracy; (b) have low complexity and scale well; (c) guarantee global monotonicity; and (d) cater for multi-class.  ...  --Agent Deep Reinforcement Learning Woojun Kim (KAIST)*; MyungSik Cho (KAIST); Youngchul Sung () 3574: Image Saliency Prediction in Transformed Domain: A Deep Complex Neural Network Method Lai Jiang (BUAA  ...  University); Ruimin Hu (Wuhan University) 2651: EA Reader: Enhance Attentive Reader for Cloze--Style Question Answering via Multi--Space Context Fusion Chengzhen Fu (Peking University)*; Yan Zhang (PEKING  ... 
doi:10.1609/aaai.v33i01.33013224 fatcat:wdto7wj635c7xdx4lyi4iwnntu

Graph Convolutional Reinforcement Learning [article]

Jiechuan Jiang, Chen Dun, Tiejun Huang, Zongqing Lu
2020 arXiv   pre-print
To tackle these difficulties, we propose graph convolutional reinforcement learning, where graph convolution adapts to the dynamics of the underlying graph of the multi-agent environment, and relation  ...  This makes it hard to learn abstract representations of mutual interplay between agents.  ...  nodes just like a neuron in a convolutional neural network (CNN).  ... 
arXiv:1810.09202v5 fatcat:lzyggbsbxva7rcgim3np6gru44

Mis-spoke or mis-lead: Achieving Robustness in Multi-Agent Communicative Reinforcement Learning [article]

Wanqi Xue, Wei Qiu, Bo An, Zinovi Rabinovich, Svetlana Obraztsova, Chai Kiat Yeo
2022 arXiv   pre-print
Recent studies in multi-agent communicative reinforcement learning (MACRL) have demonstrated that multi-agent coordination can be greatly improved by allowing communication between agents.  ...  Second, we develop a defence method based on message reconstruction, to maintain multi-agent coordination under message attacks.  ...  TarMAC [6] extends CommNet by allowing agents to pay attention to important parts of the incoming messages via Attention [37] network.  ... 
arXiv:2108.03803v2 fatcat:z5jmepxz7rf3hebs2uwvkaesxq

GUEST EDITORIAL: ARTIFICIAL INTELLIGENCE IN ENVIRONMENTAL AUTOMATION SYSTEMS

Dong Ren, Bin Li
2021 International Journal of Robotics and Automation  
(SDA) [2] ", Wang et al. address high classification accuracy for multiple types of motor imagery EEG tasks via neural network models.  ...  In "Clustering Routing Protocol Based on Game Theory in Wireless Sensor Networks for Large-scale Environmental Monitoring [1] ", Dong et al. address the imbalance energy of cluster heads via clustering  ... 
doi:10.2316/j.2021.206-0620 fatcat:d4uzuhbf65amndjs3gu2jcbsxa

Collective Intelligence for Deep Learning: A Survey of Recent Developments [article]

David Ha, Yujin Tang
2022 arXiv   pre-print
However, as these neural networks become bigger, more complex, and more widely used, fundamental problems with current deep learning models become more apparent.  ...  Advances in artificial neural networks alongside corresponding advances in hardware accelerators with large memory capacity, together with the availability of large datasets enabled practitioners to train  ...  Graph Neural Networks A class of artificial neural networks for processing data best represented by graph data structures.  ... 
arXiv:2111.14377v3 fatcat:dg5uvn7mt5g5ncgtrzw3a3ul4y
« Previous Showing results 1 — 15 out of 3,985 results