近红外无创血糖浓度的Label Sensitivity算法和支持向量机回归

近红外光谱分析技术在生物医学工程领域具有广阔应用前景。无创且持续性地测量能实时监控人体血糖水平,给糖尿病患者带来极大便利性、提高生存质量、降低糖尿病并发症发生率具有很大的社会效益。无创血糖监测的想法提出较早,但仍然存在预测精度低、预测值与标签值相关性不高等难点,至今没有达到临床要求。近年来,光谱检测技术发展迅猛且机器学习技术在智能信息处理方面具有明显优势,两者结合可以有效提高人体无创血糖医学监测模型的精度和普适性。提出了一种标签敏感度算法(LS),并结合支持向量机方法建立了人体血糖含量预测模型。使用近红外光谱仪采集了4名志愿者食指处动态血液光谱数据(每名志愿者28组数据),并使用多元散射矫正(MSC)方法消除了部分光散射的影响。考虑血糖对不同波长光的吸收有差异,提出了基于血糖浓度标签差的特征波长挑选方法,并构建了标签敏感度支持向量机(LSSVR)预测模型。设计实验,对比该模型与偏最小二乘回归(PLSR)和区分度支持向量机(FSSVR)算法。结果表明,LS算法的最佳特征波长数为32,经特征波长选择后的LSSVR表现最佳,其均方误差降低至0.02 mmol·L^(-1),明显优于全谱段PLSR模型,血糖浓度的预测值与标签值的相关系数提升至99.8%,预测值全部位于可容许误差的克拉克网格A区内。LSSVR模型的优异表现为早日实现血糖的无创监测提供了新思路。

A fast connected components labeling algorithm for binary images

A fast label-equivalence-based connected components labeling algorithm is proposed in this paper.It is a combination of two existing efficient methods,which are pivotal operations in two-pass connected components labeling algorithms.One is a fast pixel scan method,and the other is an array-based Union-Find data structure.The scan procedure assigns each foreground pixel a provisional label according to the location of the pixel.That is to say,it labels the foreground pixels following background pixels and foreground pixels in different ways,which greatly reduces the number of neighbor pixel checks.The array-based Union-Find data structure resolves the label equivalences between provisional labels by using only a single array with path compression,and it improves the efficiency of the resolving procedure which is very time-consuming in general label-equivalence-based algorithms.The experiments on various types of images with different sizes show that the proposed algorithm is superior to other labeling approaches for huge images containing many big connected components.

Novel Apriori-Based Multi-Label Learning Algorithm by Exploiting Coupled Label Relationship

It is a key challenge to exploit the label coupling relationship in multi-label classification（MLC）problems.Most previous work focused on label pairwise relations,in which generally only global statistical information is used to analyze the coupled label relationship.In this work,firstly Bayesian and hypothesis testing methods are applied to predict the label set size of testing samples within their k nearest neighbor samples,which combines global and local statistical information,and then apriori algorithm is used to mine the label coupling relationship among multiple labels rather than pairwise labels,which can exploit the label coupling relations more accurately and comprehensively.The experimental results on text,biology and audio datasets shown that,compared with the state-of-the-art algorithm,the proposed algorithm can obtain better performance on 5 common criteria.

Labeling algorithm and its fairness analysis for autonomous system

A kind of packet labeling algorithm for autonomous system is introduced. The fairness of the algorithm for each traffic stream in the integrated-services is analyzed. It is shown that the rate of each stream in the integrated-services would converge to a stable value if the transmittfing or forwarding rates converge to that of the receiving exponentially.

THE INNER-SYSTEM LABELING ALGORITHM AND ITS FAIRNESS ANALYSIS

On the basis of inner-system labeling signaling used in the integrated access system,a kind of inner-system labeling algorithm is introduced in this paper, and the fairness of the algorithm for each traffic stream in the integrated-services is analyzed. The base of this algorithm is Class of Services (CoS), and each packet entering the relative independent area (an autonomous system) would be labeled according to the service type or Quality of Service (QoS) in demand,and be scheduled and managed within the system (the system can be enlarged if conforming to the same protocol). The experimental results show that each of the stream rate in the integratedservices would converge to a stable value if the rates of transmitting converge to that of the receiving exponentially, that is, the effective traffic of each stream would be fair.

Genetic algorithm for multi-protocol label switching

A new method for multi-protocol label switching is presented in this study, whose core idea is to construct model for simulating process of accommodating network online loads and then adopt genetic algorithm to optimize the model. Due to the heuristic property of evolutional method, the new method is efficient and effective, which is verified by the experiments.

Optimization Model and Algorithm for Multi-Label Learning

This paper studies a kind of urban security risk assessment model based on multi-label learning, which is transformed into the solution of linear equations through a series of transformations, and then the solution of linear equations is transformed into an optimization problem. Finally, this paper uses some classical optimization algorithms to solve these optimization problems, the convergence of the algorithm is proved, and the advantages and disadvantages of several optimization methods are compared.

Detecting community structure using label propagation with consensus weight in complex network

Community detection is a fundamental work to analyse the structural and functional properties of complex networks. The label propagation algorithm （LPA） is a near linear time algorithm to find a good community structure. Despite various subsequent advances, an important issue of this algorithm has not yet been properly addressed. Random update orders within the algorithm severely hamper the stability of the identified community structure. In this paper, we executed the basic label propagation algorithm on networks multiple times, to obtain a set of consensus partitions. Based on these consensus partitions, we created a consensus weighted graph. In this consensus weighted graph, the weight value of the edge was the proportion value that the number of node pairs allocated in the same cluster was divided by the total number of partitions. Then, we introduced consensus weight to indicate the direction of label propagation. In label update steps, by computing the mixing value of consensus weight and label frequency, a node adopted the label which has the maximum mixing value instead of the most frequent one. For extending to different networks, we introduced a proportion parameter to adjust the proportion of consensus weight and label frequency in computing mixing value. Finally, we proposed an approach named the label propagation algorithm with consensus weight （LPAcw）, and the experimental results showed that the LPAcw could enhance considerably both the stability and the accuracy of community partitions.

Solving open vehicle problem with time window by hybrid column generation algorithm

This paper addresses the open vehicle routing problem with time window(OVRPTW), where each vehicle does not need to return to the depot after completing the delivery task.The optimization objective is to minimize the total distance. This problem exists widely in real-life logistics distribution process.We propose a hybrid column generation algorithm(HCGA) for the OVRPTW, embedding both exact algorithm and metaheuristic. In HCGA, a label setting algorithm and an intelligent algorithm are designed to select columns from small and large subproblems, respectively. Moreover, a branch strategy is devised to generate the final feasible solution for the OVRPTW. The computational results show that the proposed algorithm has faster speed and can obtain the approximate optimal solution of the problem with 100 customers in a reasonable time.

A parallel pipeline connected-component labeling method for on-orbit space target monitoring

The paper designs a peripheral maximum gray differ-ence(PMGD)image segmentation method,a connected-compo-nent labeling(CCL)algorithm based on dynamic run length(DRL),and a real-time implementation streaming processor for DRL-CCL.And it verifies the function and performance in space target monitoring scene by the carrying experiment of Tianzhou-3 cargo spacecraft(TZ-3).The PMGD image segmentation method can segment the image into highly discrete and simple point tar-gets quickly,which reduces the generation of equivalences greatly and improves the real-time performance for DRL-CCL.Through parallel pipeline design,the storage of the streaming processor is optimized by 55%with no need for external me-mory,the logic is optimized by 60%,and the energy efficiency ratio is 12 times than that of the graphics processing unit,62 times than that of the digital signal proccessing,and 147 times than that of personal computers.Analyzing the results of 8756 images completed on-orbit,the speed is up to 5.88 FPS and the target detection rate is 100%.Our algorithm and implementation method meet the requirements of lightweight,high real-time,strong robustness,full-time,and stable operation in space irradia-tion environment.

An Optimal Algorithm for Prufer Codes

This paper studies the algorithms for coding and decoding Prufer codes of a labeled tree. The algorithms for coding and decoding Prufer codes of a labeled tree in the literatures require time usually. Although there exist linear time algorithms for Prufer-like codes [1,2,3], the algorithms utilize the integer sorting algorithms. The special range of the integers to be sorted is utilized to obtain a linear time integer sorting algorithm. The Prufer code problem is reduced to integer sorting. In this paper we consider the Prufer code problem in a different angle and a more direct manner. We start from a naïve algorithm, then improved it gradually and finally we obtain a very practical linear time algorithm. The techniques we used in this paper are of interest in their own right.

面向社交网络重要信息传播的重叠节点挖掘模型研究

针对动态社交网络中的社区检测问题,提出一种面向社交网络重要信息传播的重叠节点挖掘模型(SNONMM),结合标签传播算法(LPA)和扩散激活原理,实现对动态社交网络中重叠社区的高效检测.该模型的新节点在社交网络中向其他节点传播其标签的机会大于旧节点,从而使新节点更容易被发现并纳入相应的社区.同时,引入激活值来表示每个标签的传播强度,有助于更准确地捕捉社区结构的变化.为了验证该方法的有效性,通过两个真实数据集和一个人工合成网络对其性能进行评估.实验结果表明,该方法在检测社区准确性方面优于其他可用方法.

基于特征集重构与多标签分类模型的谐波源定位方法

传统基于谐波状态估计的谐波源定位方法需要专门的同步相量量测装置,工程应用受到限制。为此,基于电能质量监测装置所采集的非同步量测数据,提出了基于特征集重构与多标签分类模型的谐波源定位方法。利用监测数据的充分统计量来挖掘量测时段的谐波信息,同时利用标签特定特征学习算法重构特征集,从而消除冗余特征以及无关特征对于谐波源定位精度的影响;提出基于邻接矩阵以及灵敏度分析的测点配置方法,结合电路网络拓扑信息实现测点的优化配置;提出基于改进极限学习机的谐波源定位方法,该方法以重构特征集为输入,建立多标签分类模型,实现谐波源定位。通过仿真与算例分析,验证了所提方法的可行性及有效性。

最省刻度尺设计的组合差集递推算法

在长度为n(n≥2为正整数)的直尺上最少刻多少个刻度就能度量1到n的所有长度,这便是至今未解决的最省刻度尺问题。阐明了最省刻度尺与极小优美图之间的关系,给出了计算最省刻度尺的所有最省刻度值的组合差集递推算法,得到长度为3~40的最省刻度尺的所有最省刻度值,同时,结合图论模型,给出了长度为41~82的最省刻度尺的最省刻度值。

基于改进MobileNet的公路行车环境安全风险源识别

为了检测公路行车环境安全风险源,为公路风险源智能化管控和随机不确定场景下实时安全风险评估提供依据,研究了基于深度卷积神经网络模型的公路行车环境安全风险源识别算法。通过改进MobileNetV3的输出层激活函数和损失函数,将网络输出的风险源类别数量由一个变为多个,解决了同一图像中存在多种风险源的识别问题。利用空间注意力机制加强MobileNetV3网络的特征提取能力,解决了MobileNetV3通道注意力机制无法关注到通道内部风险源特征信息的问题,提升了模型识别准确率。通过通道剪枝方法去除冗余扩张通道,减少网络参数量,提升了模型预测速度。实验结果表明:该方法能够有效识别行车环境安全风险源,检测率达0.829,平均分类准确率达0.833,且具备实时检测效果。

陷阱标记联合懒蚂蚁的自适应粒子群优化算法

为解决现有粒子群改进策略无法帮助已陷入局部最优和过早收敛的粒子恢复寻优性能的问题,提出一种陷阱标记联合懒蚂蚁的自适应粒子群优化(adaptive particle swarm optimization based on trap label and lazy ant, TLLA-APSO)算法。陷阱标记策略为粒子群提供动态速度增量,使其摆脱最优解的束缚。利用懒蚂蚁寻优策略多样化粒子速度,提升种群多样性。通过惯性认知策略在速度更新中引入历史位置,增加粒子的路径多样性和提升粒子的探索性能,使粒子更有效地避免陷入新的局部最优。理论证明了引入历史位置的粒子群算法的收敛性。仿真实验结果表明,所提算法不仅能有效解决粒子群已陷入局部最优和过早收敛的问题,且与其他算法相比,具有较快的收敛速度和较高的寻优精度。

双圈图的邻点可约全标号

图G(V,E)的邻点可约全标号(adjacent vertex reducible total labeling,AVRTL)是一个从V(G)∪E(G)到连续整数集{1,2,…,|V(G)|+|E(G)|}的双射,且图中所有相邻同度顶点的标号之和均相同,为S(u)=f(u)+∑uw∈E(G)f(uw).该文结合现实问题,借鉴传统遗传算法、蜂群算法等智能算法思路,设计了一种新型的AVRTL算法,通过预处理函数、调整函数等,利用循环迭代寻优的方式得到有限点内所有双圈图的邻点可约全标号结果.对实验结果进行分析,发现几类图的标号规律,总结得到若干定理并给出证明,最后给出猜想:所有的双圈图均为AVRTL图.

基于高置信度伪标签数据选择算法的临床事件抽取方法

【目的】事件抽取是构建高质量事件知识图谱的前提。临床事件抽取过程中事件元素存在依赖关系,现有方法无法准确识别事件元素并组合为事件,且现有临床事件标记数据较少,给事件抽取任务带来了极大的挑战。【方法】将临床事件抽取建模为实体识别模型,提出一种融合多特征的中文医学事件抽取方法:BERT-MCRF.该方法使用BERT构建模型的嵌入和特征提取部分,在CRF层加入多个字的滑动窗口特征,然后将BERT-MCRF作为半监督实验的基实验,提出一种高置信度伪标签数据选择算法作为筛选数据的条件,得到较高质量的300条数据与原始数据合并,最终构建了1700条语料,并重新训练模型。【结果】BERT-MCRF模型在3种属性实体上的整体F1值达到80.21%,比经典的BiLSTM-CRF模型提升15.11%;通过半监督思路重新训练的模型最终F1值达到81.56%,较原始BERT-MCRF提升了1.35%.

基于节点影响值的社区网络稳定标签传播算法

针对社区网络标签传播准确性差的问题,提出基于节点影响值的社区网络稳定标签传播算法。该算法重设社区网络相邻节点相似度,结合链路加权思想将其变换为链路无向带权图;利用节点K-shell分解值判断节点影响力,计算种子节点的中心性度量值,升序排列节点更新顺序;根据越重要节点标签影响越大的规则,修正高频数标签,获取最佳社区网络划分结果。结果表明:该算法模块度在0.320以上,准确率达到99%,因此,经该方法划分后的网络结构清晰、准确,稳定性高。

基于多标签集成学习的螺旋CT机故障诊断研究

医学应用领域计算机X线断层摄影螺旋机由于复杂的结构和较高的集成度在实际故障定位和检测中具有极高的难度;为解决这个问题,研究对螺旋CT机故障定位与检测问题进行了分析,提出一种多标签集成学习方法;该方法采用了折半查找算法获取螺旋CT机的故障数据,同时有效结合现有的卷积神经网络和循环神经网络的文本表征网络,通过自适应标签关系增强方法找出标签间的依赖关系,并利用加权约简标签集的不平衡学习能有效杜绝模型可扩展性低和模型泛化性弱等问题;经损失值、准确度、运行时间、精准率、灵敏度5个指标的实例测试结果表明,研究所给出的方法均相对于其他3种较为创新的多标签集成学习方法更具优势,且提升数值均超过2%,训练集的各个指标数据均比测试集相应数值更高;训练集和测试集中空时网络聚类约简的多标签集成学习方法的精准率分别为93.12%和87.26%,召回率分别为86.35%和84.25%;该方法能精准快速查找螺旋CT机的故障类型和故障部位,极大程度降低维修成本和延长设备的使用年限。