融合对抗训练及全局指针的实体关系联合抽取

实体关系联合抽取旨在从文本中抽取出实体关系三元组, 是构建知识图谱十分重要的步骤之一. 针对实体关系抽取中存在的信息表达能力不强、泛化能力较差、实体重叠和关系冗余等问题, 提出了一种实体关系联合抽取模型RGPNRE. 使用RoBERTa预训练模型作为编码器, 提高了模型的表达信息能力. 在训练过程中引入了对抗训练, 提升了模型的泛化能力. 使用全局指针, 解决了实体重叠的问题. 使用关系预测, 排除不可能的关系, 减少了冗余的关系. 在基于schema的中文医学信息抽取数据集CMeIE上进行的实体关系抽取实验表明, 模型的F1值比基准模型提升了约2个百分点, 在实体对重叠的情况下, 模型的F1值提升了近10个百分点, 在单一实体重叠情况下, 模型的F1值提升了大约1个百分点, 说明该模型能够更准确地提取实体关系三元组, 从而有效提升知识图谱构建的准确度. 在含有1–5个三元组的对比实验中, 在拥有4个三元组的句子中, 模型的F1值提升了约2个百分点, 而在拥有5个及以上三元组的复杂句子中, F1值提升了约1个百分点, 说明该模型能够较好地处理复杂句子场景.     

基于白鲸优化算法的配电网故障恢复方法

提出了一种基于白鲸优化（BWO）算法的配电网故障恢复方法。首先,建立光储、风储系统模型以及负荷模型,以故障后重要负荷损失量最小为目标函数,进行配电网的初步孤岛划分。其次,以网络损耗和开关操作次数加权求和最小为目标,采用BWO算法求解,获取孤岛划分与开关操作配合的故障恢复结果。然后,通过3种场景对比,验证了所提方法能够在不同故障时段获得配电网故障恢复的最优结果。最后,将BWO算法与二进制粒子群优化算法、灰狼优化算法的运行结果进行对比,验证了BWO算法寻优效果更好。     

Learning Structure from Data and Its Application to Ozone Prediction

In this paper we propose an algorithm for structure learning in predictive expert systems based on a probabilistic network representation. The idea is to have the simplest structure (minimum number of links) with acceptable predictive capability. The algorithm starts by building a tree structure based on measuring mutual information between pairs of variables, and then it adds links as necessary to obtain certain predictive performance. We have applied this method for ozone prediction in México City, where the ozone level is used as a global indicator for the air quality in different parts of the city. It is important to predict the ozone level a day, or at least several hours in advance, to reduce the health hazards and industrial losses that occur when the ozone reaches emergency levels. We obtained as a first approximation a tree-structured dependency model for predicting ozone in one part of the city. We observe that even with only three parameters, its estimations are acceptable.A causal network representation and the structure learning techniques produced some very interesting results for the ozone prediction problem. Firstly, we got some insight into the dependence structure of the phenomena. Secondly, we got an indication of which are the important and not so important variables for ozone forecasting. Taking this into account, the measurement and computational costs for ozone prediction could be reduced. And thirdly, we have obtained satisfactory short term ozone predictions based on a small set of the most important parameters.     

基于极端学习机的MANET移动性预测模型

针对移动自组织网络移动性在管理无线网络带宽资源可用性方面的重要性,为了更好地规划连续服务可用性和有效能源管理以提升网络的整体服务质量,提出了一种基于极端学习机的MANET移动性预测模型。利用ELM对MANET中的任意节点进行建模;假设已知每个移动节点当前的移动性信息（位置、速度和运动方向角度）,以这种方式预测节点未来的位置和相邻节点之间未来的距离;基于几个标准移动性模型,产生更加真实、精确的移动性预测,从而更好地捕捉任意节点直角坐标系之间现有交互/相关性。使用标准移动性模型的仿真结果验证了所提模型的有效性,实验结果表明,提出的预测模型明显改进了传统基于多层感知器的模型,此外,当预测相邻节点之间未来距离时,避免了当前算法对预测精度的限制。     

Further result on a dynamic recurrent neural-network-based adaptive observer for a class of nonlinear systems

In Kim et al. [(1997) A dynamic recurrent neural-network-based adaptive observer for a class of nonlinear systems. Automatica 33(8), 1539–1543], authors present an excellent neural network (NN) observer for a class of nonlinear systems. However, the output error equation in their paper is strictly positive real (SPR) which is restrictive assumption for nonlinear systems. In this note, by introducing a vector b₀ and Lyapunov equation, the observer design is obtained without requiring the SPR condition. Thus, our observer can be applied to a wider class of systems.     

基于BP神经网络的数据挖掘方法

神经网络的高度参数化特征使它特别灵活,以至于它可以精确的对数据进行分类,成为目前公认的高精度分类器。本文介绍一种基于BP神经网络的数据挖掘的分类方法,并提出了改进思想。     

Multiple costs based decision making with back-propagation neural networks

The current research investigates a single cost for cost-sensitive neural networks (CNN) for decision making. This may not be feasible for real cost-sensitive decisions which involve multiple costs. We propose to modify the existing model, the traditional back-propagation neural networks (TNN), by extending the back-propagation error equation for multiple cost decisions. In this multiple-cost extension, all costs are normalized to be in the same interval (i.e. between 0 and 1) as the error estimation generated in the TNN. A comparative analysis of accuracy dependent on three outcomes for constant costs was performed: (1) TNN and CNN with one constant cost (CNN-1C), (2) TNN and CNN with two constant costs (CNN-2C), and (3) CNN-1C and CNN-2C. A similar analysis for accuracy was also made for non-constant costs; (1) TNN and CNN with one non-constant cost (CNN-1NC), (2) TNN and CNN with two non-constant costs (CNN-2NC), and (3) CNN-1NC and CNN-2NC. Furthermore, we compared the misclassification cost for CNNs for both constant and non-constant costs (CNN-1C vs. CNN-2C and CNN-1NC vs. CNN-2NC). Our findings demonstrate that there is a competitive behavior between the accuracy and misclassification cost in the proposed CNN model. To obtain a higher accuracy and lower misclassification cost, our results suggest merging all constant cost matrices into one constant cost matrix for decision making. For multiple non-constant cost matrices, our results suggest maintaining separate matrices to enhance the accuracy and reduce the misclassification cost.     

多输入多输出网络系统的拥塞控制方法

该文讨论了多输入多输出网络系统中信息拥塞的控制问题,通过对造成网络拥塞的原因以及影响网络传输定性的因素进行分析,建立了解决网络拥塞的模糊逻辑控制器。仿真表明了该控制器可以有效保证网络系统中信息的稳传输。     

Consensus reaching in multi-agent packet-switched networks with non-linear coupling

An important task for multi-agent systems (MAS) is to reach a consensus, e.g. to align their velocity vectors. Recent results propose appropriate consensus protocols to achieve such tasks, but most of them do not consider the effect of communication constraints such as the presence of time-delays in the exchange of information between the agents. In this article, we provide conditions for a non-linear, locally passive MAS of arbitrary size to reach a consensus, when the agents communicate over a packet-switched network that is characterised by a given topology. Both the cases of constant and switching topologies are considered. The nature of the communication channel imposes constraints that are modelled using stochastic delays of arbitrary distribution. We first embed this model in another, distributed but deterministic delay model and provide conditions for the error introduced by this simplification. In our main result, we provide conditions for the locally passive MAS with distributed delays to reach a consensus. In the case of a fixed topology, the underlying directed graph has to contain a spanning tree. In the case of a switching topology, only the union graph of all graphs that persist over time is required to contain a spanning tree. These conditions are independent of the distribution and the size of the packet delays. To show attractivity of the consensus set, we use an invariance principle for systems described by functional differential equations based on an appropriate Lyapunov–Razumikhin function. This methodological approach is the main contribution of this work and can also be applied to other consensus problems with delays. We illustrate our results by numerical simulations showing synchronisation of non-linear Kuramoto oscillators over a digital network.     

Comparison of conventional and soft computing-based power control methods in mobile communications systems

 In this paper, we give a comparison between the conventional power control scheme and soft computing-based approaches in a mobile communications application. At the base station, the `bang–bang' control strategy and a neural network-based prediction control method are employed. In addition, full power command transmission mode, single-bit command transmission mode, and fuzzy logic-based power command enhancement unit are considered. Based on simulation experiments, we quantitatively evaluate the performance of various combinations of these control methods and command transmission modes. Conclusions on the optimal configuration are finally drawn.