Knowledge Reuse of Multi-Agent Reinforcement Learning in Cooperative Tasks

With the development and appliance of multi-agent systems, multi-agent cooperation is becoming an important problem in artificial intelligence. Multi-agent reinforcement learning (MARL) is one of the most effective methods for solving multi-agent cooperative tasks. However, the huge sample complexity of traditional reinforcement learning methods results in two kinds of training waste in MARL for cooperative tasks: all homogeneous agents are trained independently and repetitively, and multi-agent systems need training from scratch when adding a new teammate. To tackle these two problems, we propose the knowledge reuse methods of MARL. On the one hand, this paper proposes sharing experience and policy within agents to mitigate training waste. On the other hand, this paper proposes reusing the policies learned by original teams to avoid knowledge waste when adding a new agent. Experimentally, the Pursuit task demonstrates how sharing experience and policy can accelerate the training speed and enhance the performance simultaneously. Additionally, transferring the learned policies from the N-agent enables the (N+1)–agent team to immediately perform cooperative tasks successfully, and only a minor training resource can allow the multi-agents to reach optimal performance identical to that from scratch.     

An Improved Approach towards Multi-Agent Pursuit–Evasion Game Decision-Making Using Deep Reinforcement Learning

A pursuit–evasion game is a classical maneuver confrontation problem in the multi-agent systems (MASs) domain. An online decision technique based on deep reinforcement learning (DRL) was developed in this paper to address the problem of environment sensing and decision-making in pursuit–evasion games. A control-oriented framework developed from the DRL-based multi-agent deep deterministic policy gradient (MADDPG) algorithm was built to implement multi-agent cooperative decision-making to overcome the limitation of the tedious state variables required for the traditionally complicated modeling process. To address the effects of errors between a model and a real scenario, this paper introduces adversarial disturbances. It also proposes a novel adversarial attack trick and adversarial learning MADDPG (A2-MADDPG) algorithm. By introducing an adversarial attack trick for the agents themselves, uncertainties of the real world are modeled, thereby optimizing robust training. During the training process, adversarial learning was incorporated into our algorithm to preprocess the actions of multiple agents, which enabled them to properly respond to uncertain dynamic changes in MASs. Experimental results verified that the proposed approach provides superior performance and effectiveness for pursuers and evaders, and both can learn the corresponding confrontational strategy during training.     

Scalable and Transferable Reinforcement Learning for Multi-Agent Mixed Cooperative–Competitive Environments Based on Hierarchical Graph Attention

Most previous studies on multi-agent systems aim to coordinate agents to achieve a common goal, but the lack of scalability and transferability prevents them from being applied to large-scale multi-agent tasks. To deal with these limitations, we propose a deep reinforcement learning (DRL) based multi-agent coordination control method for mixed cooperative–competitive environments. To improve scalability and transferability when applying in large-scale multi-agent systems, we construct inter-agent communication and use hierarchical graph attention networks (HGAT) to process the local observations of agents and received messages from neighbors. We also adopt the gated recurrent units (GRU) to address the partial observability issue by recording historical information. The simulation results based on a cooperative task and a competitive task not only show the superiority of our method, but also indicate the scalability and transferability of our method in various scale tasks.     

Information-Theoretic Generalization Bounds for Meta-Learning and Applications

Meta-learning, or “learning to learn”, refers to techniques that infer an inductive bias from data corresponding to multiple related tasks with the goal of improving the sample efficiency for new, previously unobserved, tasks. A key performance measure for meta-learning is the meta-generalization gap, that is, the difference between the average loss measured on the meta-training data and on a new, randomly selected task. This paper presents novel information-theoretic upper bounds on the meta-generalization gap. Two broad classes of meta-learning algorithms are considered that use either separate within-task training and test sets, like model agnostic meta-learning (MAML), or joint within-task training and test sets, like reptile. Extending the existing work for conventional learning, an upper bound on the meta-generalization gap is derived for the former class that depends on the mutual information (MI) between the output of the meta-learning algorithm and its input meta-training data. For the latter, the derived bound includes an additional MI between the output of the per-task learning procedure and corresponding data set to capture within-task uncertainty. Tighter bounds are then developed for the two classes via novel individual task MI (ITMI) bounds. Applications of the derived bounds are finally discussed, including a broad class of noisy iterative algorithms for meta-learning.     

Attention-Shared Multi-Agent Actor–Critic-Based Deep Reinforcement Learning Approach for Mobile Charging Dynamic Scheduling in Wireless Rechargeable Sensor Networks

The breakthrough of wireless energy transmission (WET) technology has greatly promoted the wireless rechargeable sensor networks (WRSNs). A promising method to overcome the energy constraint problem in WRSNs is mobile charging by employing a mobile charger to charge sensors via WET. Recently, more and more studies have been conducted for mobile charging scheduling under dynamic charging environments, ignoring the consideration of the joint charging sequence scheduling and charging ratio control (JSSRC) optimal design. This paper will propose a novel attention-shared multi-agent actor–critic-based deep reinforcement learning approach for JSSRC (AMADRL-JSSRC). In AMADRL-JSSRC, we employ two heterogeneous agents named charging sequence scheduler and charging ratio controller with an independent actor network and critic network. Meanwhile, we design the reward function for them, respectively, by considering the tour length and the number of dead sensors. The AMADRL-JSSRC trains decentralized policies in multi-agent environments, using a centralized computing critic network to share an attention mechanism, and it selects relevant policy information for each agent at every charging decision. Simulation results demonstrate that the proposed AMADRL-JSSRC can efficiently prolong the lifetime of the network and reduce the number of death sensors compared with the baseline algorithms.     

Learning to Classify DWDM Optical Channels from Tiny and Imbalanced Data

Applying machine learning algorithms for assessing the transmission quality in optical networks is associated with substantial challenges. Datasets that could provide training instances tend to be small and heavily imbalanced. This requires applying imbalanced compensation techniques when using binary classification algorithms, but it also makes one-class classification, learning only from instances of the majority class, a noteworthy alternative. This work examines the utility of both these approaches using a real dataset from a Dense Wavelength Division Multiplexing network operator, gathered through the network control plane. The dataset is indeed of a very small size and contains very few examples of “bad” paths that do not deliver the required level of transmission quality. Two binary classification algorithms, random forest and extreme gradient boosting, are used in combination with two imbalance handling methods, instance weighting and synthetic minority class instance generation. Their predictive performance is compared with that of four one-class classification algorithms: One-class SVM, one-class naive Bayes classifier, isolation forest, and maximum entropy modeling. The one-class approach turns out to be clearly superior, particularly with respect to the level of classification precision, making it possible to obtain more practically useful models.     

Attention-Based Fault-Tolerant Approach for Multi-Agent Reinforcement Learning Systems

The aim of multi-agent reinforcement learning systems is to provide interacting agents with the ability to collaboratively learn and adapt to the behavior of other agents. Typically, an agent receives its private observations providing a partial view of the true state of the environment. However, in realistic settings, the harsh environment might cause one or more agents to show arbitrarily faulty or malicious behavior, which may suffice to allow the current coordination mechanisms fail. In this paper, we study a practical scenario of multi-agent reinforcement learning systems considering the security issues in the presence of agents with arbitrarily faulty or malicious behavior. The previous state-of-the-art work that coped with extremely noisy environments was designed on the basis that the noise intensity in the environment was known in advance. However, when the noise intensity changes, the existing method has to adjust the configuration of the model to learn in new environments, which limits the practical applications. To overcome these difficulties, we present an Attention-based Fault-Tolerant (FT-Attn) model, which can select not only correct, but also relevant information for each agent at every time step in noisy environments. The multihead attention mechanism enables the agents to learn effective communication policies through experience concurrent with the action policies. Empirical results showed that FT-Attn beats previous state-of-the-art methods in some extremely noisy environments in both cooperative and competitive scenarios, much closer to the upper-bound performance. Furthermore, FT-Attn maintains a more general fault tolerance ability and does not rely on the prior knowledge about the noise intensity of the environment.     

Explainable AI: A Review of Machine Learning Interpretability Methods

Recent advances in artificial intelligence (AI) have led to its widespread industrial adoption, with machine learning systems demonstrating superhuman performance in a significant number of tasks. However, this surge in performance, has often been achieved through increased model complexity, turning such systems into “black box” approaches and causing uncertainty regarding the way they operate and, ultimately, the way that they come to decisions. This ambiguity has made it problematic for machine learning systems to be adopted in sensitive yet critical domains, where their value could be immense, such as healthcare. As a result, scientific interest in the field of Explainable Artificial Intelligence (XAI), a field that is concerned with the development of new methods that explain and interpret machine learning models, has been tremendously reignited over recent years. This study focuses on machine learning interpretability methods; more specifically, a literature review and taxonomy of these methods are presented, as well as links to their programming implementations, in the hope that this survey would serve as a reference point for both theorists and practitioners.     

Co-Evolution of Predator-Prey Ecosystems by Reinforcement Learning Agents

The problem of finding adequate population models in ecology is important for understanding essential aspects of their dynamic nature. Since analyzing and accurately predicting the intelligent adaptation of multiple species is difficult due to their complex interactions, the study of population dynamics still remains a challenging task in computational biology. In this paper, we use a modern deep reinforcement learning (RL) approach to explore a new avenue for understanding predator-prey ecosystems. Recently, reinforcement learning methods have achieved impressive results in areas, such as games and robotics. RL agents generally focus on building strategies for taking actions in an environment in order to maximize their expected returns. Here we frame the co-evolution of predators and preys in an ecosystem as allowing agents to learn and evolve toward better ones in a manner appropriate for multi-agent reinforcement learning. Recent significant advancements in reinforcement learning allow for new perspectives on these types of ecological issues. Our simulation results show that throughout the scenarios with RL agents, predators can achieve a reasonable level of sustainability, along with their preys.     

Estimating the Individual Treatment Effect on Survival Time Based on Prior Knowledge and Counterfactual Prediction

The estimation of the Individual Treatment Effect (ITE) on survival time is an important research topic in clinics-based causal inference. Various representation learning methods have been proposed to deal with its three key problems, i.e., reducing selection bias, handling censored survival data, and avoiding balancing non-confounders. However, none of them consider all three problems in a single method. In this study, by combining the Counterfactual Survival Analysis (CSA) model and Dragonnet from the literature, we first propose a CSA–Dragonnet to deal with the three problems simultaneously. Moreover, we found that conclusions from traditional Randomized Controlled Trials (RCTs) or Retrospective Cohort Studies (RCSs) can offer valuable bound information to the counterfactual learning of ITE, which has never been used by existing ITE estimation methods. Hence, we further propose a CSA–Dragonnet with Embedded Prior Knowledge (CDNEPK) by formulating a unified expression of the prior knowledge given by RCTs or RCSs, inserting counterfactual prediction nets into CSA–Dragonnet and defining loss items based on the bounds for the ITE extracted from prior knowledge. Semi-synthetic data experiments showed that CDNEPK has superior performance. Real-world experiments indicated that CDNEPK can offer meaningful treatment advice.     

TNT: An Interpretable Tree-Network-Tree Learning Framework using Knowledge Distillation

Deep Neural Networks (DNNs) usually work in an end-to-end manner. This makes the trained DNNs easy to use, but they remain an ambiguous decision process for every test case. Unfortunately, the interpretability of decisions is crucial in some scenarios, such as medical or financial data mining and decision-making. In this paper, we propose a Tree-Network-Tree (TNT) learning framework for explainable decision-making, where the knowledge is alternately transferred between the tree model and DNNs. Specifically, the proposed TNT learning framework exerts the advantages of different models at different stages: (1) a novel James–Stein Decision Tree (JSDT) is proposed to generate better knowledge representations for DNNs, especially when the input data are in low-frequency or low-quality; (2) the DNNs output high-performing prediction result from the knowledge embedding inputs and behave as a teacher model for the following tree model; and (3) a novel distillable Gradient Boosted Decision Tree (dGBDT) is proposed to learn interpretable trees from the soft labels and make a comparable prediction as DNNs do. Extensive experiments on various machine learning tasks demonstrated the effectiveness of the proposed method.     

FTRLIM: Distributed Instance Matching Framework for Large-Scale Knowledge Graph Fusion

Instance matching is a key task in knowledge graph fusion, and it is critical to improving the efficiency of instance matching, given the increasing scale of knowledge graphs. Blocking algorithms selecting candidate instance pairs for comparison is one of the effective methods to achieve the goal. In this paper, we propose a novel blocking algorithm named MultiObJ, which constructs indexes for instances based on the Ordered Joint of Multiple Objects’ features to limit the number of candidate instance pairs. Based on MultiObJ, we further propose a distributed framework named Follow-the-Regular-Leader Instance Matching (FTRLIM), which matches instances between large-scale knowledge graphs with approximately linear time complexity. FTRLIM has participated in OAEI 2019 and achieved the best matching quality with significantly efficiency. In this research, we construct three data collections based on a real-world large-scale knowledge graph. Experiment results on the constructed data collections and two real-world datasets indicate that MultiObJ and FTRLIM outperform other state-of-the-art methods.     

促知识内化的学习活动设计研究

常规物理学习模式中,教师理应从学生实际需求出发,多元化授教,引导学生构建有效的认知网络,并在此基础上促进学生知识内化.经远程问卷"物理课程学生知识内化"调查,梳理学习中存在的问题,提出"促知识内化的学习活动模式设计",并进行实证研究.以此提出促知识内化的学习活动设计对策与策略.一是巧设演示实验进行情景导入,激发学生兴趣;二是依托知识图谱技术,帮助学生完成新旧知识的连接,促进其知识结构的补齐;三是布置"进化型"作业,以跨界、多元、有趣和思维表达为特点,辅以多元评价、精准反馈、持续激励为手段.     

Information Flows of Diverse Autoencoders

Deep learning methods have had outstanding performances in various fields. A fundamental query is why they are so effective. Information theory provides a potential answer by interpreting the learning process as the information transmission and compression of data. The information flows can be visualized on the information plane of the mutual information among the input, hidden, and output layers. In this study, we examine how the information flows are shaped by the network parameters, such as depth, sparsity, weight constraints, and hidden representations. Here, we adopt autoencoders as models of deep learning, because (i) they have clear guidelines for their information flows, and (ii) they have various species, such as vanilla, sparse, tied, variational, and label autoencoders. We measured their information flows using Rényi’s matrix-based α-order entropy functional. As learning progresses, they show a typical fitting phase where the amounts of input-to-hidden and hidden-to-output mutual information both increase. In the last stage of learning, however, some autoencoders show a simplifying phase, previously called the “compression phase”, where input-to-hidden mutual information diminishes. In particular, the sparsity regularization of hidden activities amplifies the simplifying phase. However, tied, variational, and label autoencoders do not have a simplifying phase. Nevertheless, all autoencoders have similar reconstruction errors for training and test data. Thus, the simplifying phase does not seem to be necessary for the generalization of learning.     

Space-Air-Ground Integrated Mobile Crowdsensing for Partially Observable Data Collection by Multi-Scale Convolutional Graph Reinforcement Learning

Mobile crowdsensing (MCS) is attracting considerable attention in the past few years as a new paradigm for large-scale information sensing. Unmanned aerial vehicles (UAVs) have played a significant role in MCS tasks and served as crucial nodes in the newly-proposed space-air-ground integrated network (SAGIN). In this paper, we incorporate SAGIN into MCS task and present a Space-Air-Ground integrated Mobile CrowdSensing (SAG-MCS) problem. Based on multi-source observations from embedded sensors and satellites, an aerial UAV swarm is required to carry out energy-efficient data collection and recharging tasks. Up to date, few studies have explored such multi-task MCS problem with the cooperation of UAV swarm and satellites. To address this multi-agent problem, we propose a novel deep reinforcement learning (DRL) based method called Multi-Scale Soft Deep Recurrent Graph Network (ms-SDRGN). Our ms-SDRGN approach incorporates a multi-scale convolutional encoder to process multi-source raw observations for better feature exploitation. We also use a graph attention mechanism to model inter-UAV communications and aggregate extra neighboring information, and utilize a gated recurrent unit for long-term performance. In addition, a stochastic policy can be learned through a maximum-entropy method with an adjustable temperature parameter. Specifically, we design a heuristic reward function to encourage the agents to achieve global cooperation under partial observability. We train the model to convergence and conduct a series of case studies. Evaluation results show statistical significance and that ms-SDRGN outperforms three state-of-the-art DRL baselines in SAG-MCS. Compared with the best-performing baseline, ms-SDRGN improves 29.0% reward and 3.8% CFE score. We also investigate the scalability and robustness of ms-SDRGN towards DRL environments with diverse observation scales or demanding communication conditions.     

Bottleneck Problems: An Information and Estimation-Theoretic View

Information bottleneck (IB) and privacy funnel (PF) are two closely related optimization problems which have found applications in machine learning, design of privacy algorithms, capacity problems (e.g., Mrs. Gerber’s Lemma), and strong data processing inequalities, among others. In this work, we first investigate the functional properties of IB and PF through a unified theoretical framework. We then connect them to three information-theoretic coding problems, namely hypothesis testing against independence, noisy source coding, and dependence dilution. Leveraging these connections, we prove a new cardinality bound on the auxiliary variable in IB, making its computation more tractable for discrete random variables. In the second part, we introduce a general family of optimization problems, termed “bottleneck problems”, by replacing mutual information in IB and PF with other notions of mutual information, namely f-information and Arimoto’s mutual information. We then argue that, unlike IB and PF, these problems lead to easily interpretable guarantees in a variety of inference tasks with statistical constraints on accuracy and privacy. While the underlying optimization problems are non-convex, we develop a technique to evaluate bottleneck problems in closed form by equivalently expressing them in terms of lower convex or upper concave envelope of certain functions. By applying this technique to a binary case, we derive closed form expressions for several bottleneck problems.     

Deep transfer reinforcement learning for resource allocation in hybrid multiple access systems

This paper proposes a resource allocation scheme for hybrid multiple access involving both orthogonal multiple access and non-orthogonal multiple access (NOMA) techniques. The proposed resource allocation scheme employs multi-agent deep reinforcement learning (MA-DRL) to maximize the sum-rate for all users. More specifically, the MA-DRL-based scheme jointly allocates subcarrier and power resources for users by utilizing deep Q networks and multi-agent deep deterministic policy gradient networks. Meanwhile, an adaptive learning determiner mechanism is introduced into our allocation scheme to achieve better sum-rate performance. However, the above deep reinforcement learning adopted by our scheme cannot optimize parameters quickly in the new communication model. In order to better adapt to the new environment and make the resource allocation strategy more robust, we propose a transfer learning scheme based on deep reinforcement learning (T-DRL). The T-DRL-based scheme allows us to transfer the subcarrier allocation network and the power allocation network collectively or independently. Simulation results show that the proposed MA-DRL-based resource allocation scheme can achieve better sum-rate performance. Furthermore, the T-DRL-based scheme can effectively improve the convergence speed of the deep resource allocation network.     

An Information Theoretic Approach to Symbolic Learning in Synthetic Languages

An important aspect of using entropy-based models and proposed “synthetic languages”, is the seemingly simple task of knowing how to identify the probabilistic symbols. If the system has discrete features, then this task may be trivial; however, for observed analog behaviors described by continuous values, this raises the question of how we should determine such symbols. This task of symbolization extends the concept of scalar and vector quantization to consider explicit linguistic properties. Unlike previous quantization algorithms where the aim is primarily data compression and fidelity, the goal in this case is to produce a symbolic output sequence which incorporates some linguistic properties and hence is useful in forming language-based models. Hence, in this paper, we present methods for symbolization which take into account such properties in the form of probabilistic constraints. In particular, we propose new symbolization algorithms which constrain the symbols to have a Zipf–Mandelbrot–Li distribution which approximates the behavior of language elements. We introduce a novel constrained EM algorithm which is shown to effectively learn to produce symbols which approximate a Zipfian distribution. We demonstrate the efficacy of the proposed approaches on some examples using real world data in different tasks, including the translation of animal behavior into a possible human language understandable equivalent.     

Study of Nonlinear Models of Oscillatory Systems by Applying an Intelligent Computational Technique

In this paper, we have analyzed the mathematical model of various nonlinear oscillators arising in different fields of engineering. Further, approximate solutions for different variations in oscillators are studied by using feedforward neural networks (NNs) based on the backpropagated Levenberg–Marquardt algorithm (BLMA). A data set for different problem scenarios for the supervised learning of BLMA has been generated by the Runge–Kutta method of order 4 (RK-4) with the “NDSolve” package in Mathematica. The worth of the approximate solution by NN-BLMA is attained by employing the processing of testing, training, and validation of the reference data set. For each model, convergence analysis, error histograms, regression analysis, and curve fitting are considered to study the robustness and accuracy of the design scheme.