Large neighborhood search for the bike request scheduling problem

This paper introduces an efficient algorithm for the bike request scheduling problem (BRSP). The BRSP is built around the concept of request, defined as the pickup or dropoff of a number of identical items (bikes) at a specific station, within a certain time window, and with a certain priority. The aim of the BRSP is to sequence requests on (and hence determine the routes of) a set of vehicles, in such a way that the sum of the priorities of the executed requests is maximized, all time windows are respected, and the capacity of the vehicles is not exceeded. The generation of the set of requests is explicitly not a part of the problem definition of the BRSP. The primary application of the BRSP, from which it derives its name, is to determine the routes of a set of repositioning vehicles in a bike sharing system, although other applications exist. The algorithm introduced in this paper is based on a set of related greedy randomized adaptive search procedure followed by variable neighborhood descent (GRASP + VND) operators embedded in a large neighborhood search (LNS) framework. Since this paper presents the first heuristic for the BRSP, a computational comparison to existing approaches is not possible. We therefore compare the solutions found by our LNS heuristic to those found by an exact solver (Gurobi). These experiments confirm that the proposed algorithm scales to realistic dimensions and is able to find near‐optimal solutions in seconds.     

R2D2: A scalable deep learning toolkit for medical imaging segmentation

Deep learning has gained a significant popularity in recent years thanks to its tremendous success across a wide range of relevant fields of applications, including medical image analysis domain in particular. Although convolutional neural networks (CNNs) based medical applications have been providing powerful solutions and revolutionizing medicine, efficiently training of CNNs models is a tedious and challenging task. It is a computationally intensive process taking long time and rare system resources, which represents a significant hindrance to scientific research progress. In order to address this challenge, we propose in this article, R2D2, a scalable intuitive deep learning toolkit for medical imaging semantic segmentation. To the best of our knowledge, the present work is the first that aims to tackle this issue by offering a novel distributed versions of two well-known and widely used CNN segmentation architectures [ie, fully convolutional network (FCN) and U-Net]. We introduce the design and the core building blocks of R2D2. We further present and analyze its experimental evaluation results on two different concrete medical imaging segmentation use cases. R2D2 achieves up to 17.5× and 10.4× speedup than single-node based training of U-Net and FCN, respectively, with a negligible, though still unexpected segmentation accuracy loss. R2D2 offers not only an empirical evidence and investigates in-depth the latest published works but also it facilitates and significantly reduces the effort required by researchers to quickly prototype and easily discover cutting-edge CNN configurations and architectures.     

A benchmark data set for two-phase Coriolis metering

For more than a decade there has been growing interest in the use of Coriolis mass flow metering applied to two-phase (gas/liquid) and multiphase (oil/water/gas) conditions. It is well-established that the mass flow and density measurements generated from multiphase flows are subject to large errors, and a variety of physical models and correction techniques have been proposed to explain and/or to compensate for these errors. One difficulty is the absence of a common basis for comparing correction techniques, because different flowtube designs and configurations, as well as liquid and gas properties, may result in quite different error curves. Furthermore, some researchers with interests in the modelling aspects of the field may not have suitable multiphase laboratory facilities to generate their own data sets. This paper offers a small data set that may be used by researchers as a benchmark i.e. a common data set for comparing correction techniques. The data set was collected at the UK National Flow Laboratory TUV-NEL, using air and a viscous oil, and provides experimental points over a wide flow range (8:1 turndown) and with Gas Volume Fraction (GVF) values up to 60%. As a first investigation using the benchmark data set, we consider how data sparsity (i.e. the flow rate and GVF spacing in the experimental grid) affects the accuracy of a correction model. A range of neural network models are evaluated, based on different subsets of the benchmark data set. The data set and some exemplary code are provided with the paper. Additional data sets are available on a web site created to support this initiative.     

伪标签置信选择的半监督集成学习视频语义检测

在视频语义检测中，有标记样本不足会严重影响检测的性能，而且伪标签样本中的噪声也会导致集成学习基分类器性能提升不足。为此，提出一种伪标签置信选择的半监督集成学习算法。首先，在三个不同的特征空间上训练出三个基分类器，得到基分类器的标签矢量；然后，引入加权融合样本所属某个类别的最大概率与次大概率的误差和样本所属某个类别的最大概率与样本所属其他各类别的平均概率的误差，作为基分类器的标签置信度，并融合标签矢量和标签置信度得到样本的伪标签和集成置信度；接着，选择集成置信度高的样本加入到有标签的样本集，迭代训练基分类器；最后，采用训练好的基分类器集成协作检测视频语义概念。该算法在实验数据集UCF11上的平均准确率到达了83.48%，与Co-KNN-SVM算法相比，平均准确率提高了3.48个百分点。该算法选择的伪标签能体现样本所属类别与其他类别的总体差异性，又能体现所属类别的唯一性，可减少利用伪标签样本的风险，有效提高视频语义概念检测的准确率。     

Title加TextRank抽取关键句的情感分类研究

考虑到不同句子对判断文档情感倾向的重要程度不同，因而区分文档的关键句和细节句将有助于提高情感分类的性能。同时，考虑到Title和上下文信息，提出了一种基于Title和加权TextRank抽取关键句的情感分析方法SKTT，实现了高效的情感分析。根据文档Title的情感权重计算Title贡献度，考虑到标点和语义规则对情感倾向的影响；根据加权TextRank算法思想，在文档正文中构建了一个情感句有向图来提取关键句；计算所有关键句的情感倾向进行情感分类。在4个领域上进行实验，实验结果表明，该SKTT方法性能明显优于Baseline，具有高效性。     

基于变异算子和邻域值自适应的MOEA/D算法

基于分解的多目标进化算法（MOEA/D）在解决多目标问题时，具有简单有效的特点。但多数MOEA/D采用固定的控制参数，导致全局搜索能力差，难以平衡收敛性和多样性。针对以上问题提出一种基于变异算子和邻域值自适应的多目标优化算法。该算法根据种群中个体适应度值的分散或集中程度进行判断，并据此对变异算子进行自适应的调节，从而增强算法的全局搜索能力；根据进化所处的阶段以及个体适应度值的集中程度，自适应地调节邻域值大小，保证每个个体在不同的进化代数都有一个邻域值大小；在子问题邻域中，统计子问题对应个体的被支配数，通过判断被支配数是否超过设定的上限，来决定是否将Pareto支配关系也作为邻域内判断个体好坏的准则之一。将提出的算法与传统的MOEA/D在标准测试问题上进行对比。实验结果表明，提出的算法求得的解集具有更好的收敛性和多样性，在求解性能上具有一定的优势。     

基于改进粗糙集理论的SAR图像目标增强

图像增强在提高SAR图像舰船目标检测精度方面具有十分重要的意义。由于传统算法不能很好地对SAR图像进行目标增强，提出了基于改进粗糙集理论和引力场强度的目标增强算法。通过借鉴引力场相关理论知识，将粗糙集条件属性集中的梯度属性改进为引力场强度属性，从而实现对原图像的目标增强。与其他算法进行了实验比较，结果表明提出的改进算法相比于其他算法更适用于SAR图像特性，能更好地对舰船目标像素进行针对性增强，具备一定的工程应用价值。     

一种有效折扣证据源的冲突证据合成方法

为解决证据组合规则中一票否决和弱决策证据在低冲突情况下出现的反直观推理的现象，提出一种有效折扣证据源的冲突证据合成方法。首先根据证据的信任函数和似然函数求得证据间的相似度，然后求出证据之间的支持程度，并确定折扣因子，最后采用Dempster组合规则合成利用折扣因子修正后的证据源。数值算例分析结果表明，改进后的证据组合方法可以有效地处理证据冲突。     

基于聚类的数据加权优化在犯罪预测中的应用

近年来，我国传统暴力犯罪与成年人犯罪呈下降态势，但是，犯罪案由层出不穷。为有效提升公安实践工作中犯罪预测能力，打击各类违法犯罪事件,本文针对犯罪数据，提出一种新型犯罪预测模型。利用密度聚类分析方法将犯罪数据分类，然后进行数据降维提取关键属性生成特征数据，继而对特征数据进行加权优化并采用机器学习的方式对特征数据进行学习，从而预测犯罪案由。实验结果表明，与传统方法相比，本文方法具有更好的预测效果，为公安实践工作中类似案件的侦破和预防，提供新的路径支撑。     

物联网环境下基于情境的语义Web服务选择

当前Web服务海量增加，物联网应用技术快速发展、不断普及，而现有的Web服务选择算法低效、用户匹配度低。针对该问题提出一种物联网环境下基于情境的语义Web服务选择方法。该方法应用QoS参数的无量纲化与语义Web服务动态选择方式，将物联网环境下服务与语义Web服务相结合，并根据用户需求针对QoS选择最优的服务集。实验表明，该方法能有效地提高用户服务动态选择的成功率。