arXiv subjects

Shih-wei Liao

Publications and source records attributed to Shih-wei Liao.

6 recordsLinked to original sources

arXiv2023

Rethinking the Implementation Tricks and Monotonicity Constraint in Cooperative Multi-Agent Reinforcement Learning

Many complex multi-agent systems such as robot swarms control and autonomous vehicle coordination can be modeled as Multi-Agent Reinforcement Learning (MARL) tasks. QMIX, a widely popular MARL algorithm, has been used as a baseline for the benchmark environments, e.g., Starcraft Multi-Agent Challenge (SMAC), Difficulty-Enhanced Predator-Prey (DEPP). Recent variants of QMIX target relaxing the monotonicity constraint of QMIX, allowing for performance improvement in SMAC. In this paper, we investigate the code-level optimizations of these variants and the monotonicity constraint. (1) We find that such improvements of the variants are significantly affected by various code-level optimizations. (2) The experiment results show that QMIX with normalized optimizations outperforms other works in SMAC; (3) beyond the common wisdom from these works, the monotonicity constraint can improve sample efficiency in SMAC and DEPP. We also discuss why monotonicity constraints work well in purely cooperative tasks with a theoretical analysis. We open-source the code at \url{https://github.com/hijkzzz/pymarl2}.

cs.LG↗

arXiv2023

Policy Regularization via Noisy Advantage Values for Cooperative Multi-agent Actor-Critic methods

Recent works have applied the Proximal Policy Optimization (PPO) to the multi-agent cooperative tasks, such as Independent PPO (IPPO); and vanilla Multi-agent PPO (MAPPO) which has a centralized value function. However, previous literature shows that MAPPO may not perform as well as Independent PPO (IPPO) and the Fine-tuned QMIX on Starcraft Multi-Agent Challenge (SMAC). MAPPO-Feature-Pruned (MAPPO-FP) improves the performance of MAPPO by the carefully designed agent-specific features, which may be not friendly to algorithmic utility. By contrast, we find that MAPPO may face the problem of \textit{The Policies Overfitting in Multi-agent Cooperation(POMAC)}, as they learn policies by the sampled advantage values. Then POMAC may lead to updating the multi-agent policies in a suboptimal direction and prevent the agents from exploring better trajectories. In this paper, to mitigate the multi-agent policies overfitting, we propose a novel policy regularization method, which disturbs the advantage values via random Gaussian noise. The experimental results show that our method outperforms the Fine-tuned QMIX, MAPPO-FP, and achieves SOTA on SMAC without agent-specific features. We open-source the code at \url{https://github.com/hijkzzz/noisy-mappo}.

cs.MA↗

arXiv2021

QR-MIX: Distributional Value Function Factorisation for Cooperative Multi-Agent Reinforcement Learning

In Cooperative Multi-Agent Reinforcement Learning (MARL) and under the setting of Centralized Training with Decentralized Execution (CTDE), agents observe and interact with their environment locally and independently. With local observation and random sampling, the randomness in rewards and observations leads to randomness in long-term returns. Existing methods such as Value Decomposition Network (VDN) and QMIX estimate the value of long-term returns as a scalar that does not contain the information of randomness. Our proposed model QR-MIX introduces quantile regression, modeling joint state-action values as a distribution, combining QMIX with Implicit Quantile Network (IQN). However, the monotonicity in QMIX limits the expression of joint state-action value distribution and may lead to incorrect estimation results in non-monotonic cases. Therefore, we proposed a flexible loss function to approximate the monotonicity found in QMIX. Our model is not only more tolerant of the randomness of returns, but also more tolerant of the randomness of monotonic constraints. The experimental results demonstrate that QR-MIX outperforms the previous state-of-the-art method QMIX in the StarCraft Multi-Agent Challenge (SMAC) environment.

cs.LG↗

arXiv2020

An Uncoupled Training Architecture for Large Graph Learning

Graph Convolutional Network (GCN) has been widely used in graph learning tasks. However, GCN-based models (GCNs) is an inherently coupled training framework repetitively conducting the complex neighboring aggregation, which leads to the limitation of flexibility in processing large-scale graph. With the depth of layers increases, the computational and memory cost of GCNs grow explosively due to the recursive neighborhood expansion. To tackle these issues, we present Node2Grids, a flexible uncoupled training framework that leverages the independent mapped data for obtaining the embedding. Instead of directly processing the coupled nodes as GCNs, Node2Grids supports a more efficacious method in practice, mapping the coupled graph data into the independent grid-like data which can be fed into the efficient Convolutional Neural Network (CNN). This simple but valid strategy significantly saves memory and computational resource while achieving comparable results with the leading GCN-based models. Specifically, by ranking each node's influence through degree, Node2Grids selects the most influential first-order as well as second-order neighbors with central node fusion information to construct the grid-like data. For further improving the efficiency of downstream tasks, a simple CNN-based neural network is employed to capture the significant information from the mapped grid-like data. Moreover, the grid-level attention mechanism is implemented, which enables implicitly specifying the different weights for neighboring nodes with different influences. In addition to the typical transductive and inductive learning tasks, we also verify our framework on million-scale graphs to demonstrate the superiority of the proposed Node2Grids model against the state-of-the-art GCN-based approaches.

cs.LG↗

arXiv2019

An Evaluation of Bitcoin Address Classification based on Transaction History Summarization

Bitcoin is a cryptocurrency that features a distributed, decentralized and trustworthy mechanism, which has made Bitcoin a popular global transaction platform. The transaction efficiency among nations and the privacy benefiting from address anonymity of the Bitcoin network have attracted many activities such as payments, investments, gambling, and even money laundering in the past decade. Unfortunately, some criminal behaviors which took advantage of this platform were not identified. This has discouraged many governments to support cryptocurrency. Thus, the capability to identify criminal addresses becomes an important issue in the cryptocurrency network. In this paper, we propose new features in addition to those commonly used in the literature to build a classification model for detecting abnormality of Bitcoin network addresses. These features include various high orders of moments of transaction time (represented by block height) which summarizes the transaction history in an efficient way. The extracted features are trained by supervised machine learning methods on a labeling category data set. The experimental evaluation shows that these features have improved the performance of Bitcoin address classification significantly. We evaluate the results under eight classifiers and achieve the highest Micro-F1/Macro-F1 of 87%/86% with LightGBM.

cs.CR↗

arXiv2018

Personalized Difficulty Adjustment for Countering the Double-Spending Attack in Proof-of-Work Consensus Protocols

Bitcoin is the first secure decentralized electronic currency system. However, it is known to be inefficient due to its proof-of-work (PoW) consensus algorithm and has the potential hazard of double spending. In this paper, we aim to reduce the probability of double spending by decreasing the probability of consecutive winning. We first formalize a PoW-based decentralized secure network model in order to present a quantitative analysis. Next, to resolve the risk of double spending, we propose the personalized difficulty adjustment (PDA) mechanism which modifies the difficulty of each participant such that those who win more blocks in the past few rounds have a smaller probability to win in the next round. To analyze the performance of the PDA mechanism, we observe that the system can be modeled by a high-order Markov chain. Finally, we show that PDA effectively decreases the probability of consecutive winning and results in a more trustworthy PoW-based system.

cs.CR↗