Genetic algorithms in classifier fusion

An intense research around classifier fusion in recent years revealed that combining performance strongly depends on careful selection of classifiers to be combined. Classifier performance depends, in turn, on careful selection of features, which could be further restricted by the subspaces of the data domain. On the other hand, there is already a number of classifier fusion techniques available and the choice of the most suitable method depends back on the selections made within classifier, features and data spaces. In all these multidimensional selection tasks genetic algorithms (GA) appear to be one of the most suitable techniques providing reasonable balance between searching complexity and the performance of the solutions found. In this work, an attempt is made to revise the capability of genetic algorithms to be applied to selection across many dimensions of the classifier fusion process including data, features, classifiers and even classifier combiners. In the first of the discussed models the potential for combined classification improvement by GA-selected weights for the soft combining of classifier outputs has been investigated. The second of the proposed models describes a more general system where the specifically designed GA is applied to selection carried out simultaneously along many dimensions of the classifier fusion process. Both, the weighted soft combiners and the prototype of the three-dimensional fusion–classifier–feature selection model have been developed and tested using typical benchmark datasets and some comparative experimental results are also presented.     

基于多分类器融合的步态识别方法

为了提高现有基于智能手机加速度传感器步态身份识别的性能,提出了一种基于多分类器融合（MCF）的识别方法。首先,针对现有方法所提取的步态特征较为单一的问题,对单个步态周期提取相对匀变加速度的速度变化量,以及单位时间内加速度变化量作为两类新特征（共16个）;其次,将新特征结合常用的时域、频域特征组成新的特征集,用于训练识别效果与训练时间俱佳的多个分类器;最后,采用多尺度投票法（MSV）对多分类器的输出进行融合处理,得到最终的分类结果。为了检测该方法的性能,采集了32个志愿者的步态数据。实验结果表明,新特征对于单个分类器的识别率平均提升5.95个百分点,最终通过MSV融合算法的识别率为97.78%。     

基于依存句法分析与分类器融合的触发词抽取方法

事件检测与分类是事件抽取的关键环节,触发词抽取是完成事件检测与分类的主流方法。提出了一种事件触发词抽取方法,该方法针对单一触发词抽取方法没有充分利用依存句法分析信息且召回率不高的问题,通过综合利用依存句法分析信息和其他信息抽取触发词-实体描述对的方法来提高触发词抽取的召回率,然后将触发词-实体描述对抽取结果与单一触发词抽取结果相融合以避免召回率提高所带来的准确率下降问题。在ACE2005中文语料上进行实验,该方法在事件检测与分类任务中取得较好效果,F值分别达到了69.0%和66.2%。     

Controlling the diversity in classifier ensembles through a measure of agreement

In this paper, a simulation method is proposed to generate a set of classifier outputs with specified individual accuracies and fixed pairwise agreement. A diversity measure (kappa) is used to control the agreement among classifiers for building the classifier teams. The generated team outputs can be used to study the behaviour of class-type combination methods such as voting rules over multiple dependent classifiers.     

On the impact of fusion strategies on classification errors for large ensembles of classifiers

The growing availability of sensor networks brings practical situations where a large number of classifiers can be used for building a classifier ensemble. In the most general case involving sensor networks, the classifiers are fed with multiple inputs collected at different locations. However, classifier fusion is often studied within an idealized formulation where each classifier is fed with the same point in the feature space, and estimate the posterior class probability given this input. We first expand this formulation to situations where classifiers are fed with multiple inputs, demonstrating the relevance of the formulation to situations involving sensor networks, and a large number of classifiers. Following that, we determine the rate of convergence of the classification error of a classifier ensemble for three fusion strategies (average, median and maximum) when the number of classifiers becomes large. As the size of the ensemble increases, the best strategy is defined as the one that results in fastest convergence of the classification error to zero. The best strategy is analytically shown to depend on the distribution of the individual classification errors: average is the best for normal distributions; maximum is the best for uniform distributions; and median is the best for Cauchy distributions. The general effect of heavy-tailedness is also analytically investigated for the average and median strategies. The median strategy is shown to be robust to heavy-tailedness, while performance of the average strategy is shown to degrade as heavy-tailedness becomes more pronounced. The combined effects of bimodality and heavy-tailedness are also investigated when the number of classifiers become large.     

On output independence and complementariness in rank-based multiple classifier decision systems

This study presents a theoretical analysis of output independence and complementariness between classifiers in a rank-based multiple classifier decision system in the context of the partitioned observation space theory. To enable such an analysis, an information theoretic interpretation of a rank-based multiple classifier system is developed and basic concepts from information theory are applied to develop measures for output independence and complementariness. It is shown that output independence of classifiers is not a requirement for achieving complementariness between these classifiers. Namely, output independence does not imply a performance improvement by combining multiple classifiers. A condition called dominance is shown to be important instead. The information theoretic measures proposed for output independence and complementariness are justified by simulated examples.     

An investigation of the effects of correlation and autocorrelation on classifier fusion and optimal classifier ensembles

The purpose of this research was to study various fusion strategies where the levels of correlation between features and auto-correlation within features could be controlled. The fusion strategies were chosen to reflect decision-level fusion (ISOC and ROC), feature level fusion, via a single Generalized Regression Neural Network (GRNN) employing all available features, and an intermediate level of fusion that employed the outputs of individual classifiers, in this case posterior probability estimates, before they are subjected to thresholds and mapped into decisions. This latter scheme involved fusing the posterior probability estimates by employing them as features in a probabilistic neural network. Correlation was injected into the data set both within a feature set (auto-correlation) and across feature sets, and sample size was varied for a two class problem. The fusion methods were then extended to three classifiers, and a method is demonstrated that selects the optimal classifier ensemble.     

Measures of Diversity in Classifier Ensembles and Their Relationship with the Ensemble Accuracy

Diversity among the members of a team of classifiers is deemed to be a key issue in classifier combination. However, measuring diversity is not straightforward because there is no generally accepted formal definition. We have found and studied ten statistics which can measure diversity among binary classifier outputs (correct or incorrect vote for the class label): four averaged pairwise measures (the Q statistic, the correlation, the disagreement and the double fault) and six non-pairwise measures (the entropy of the votes, the difficulty index, the Kohavi-Wolpert variance, the interrater agreement, the generalized diversity, and the coincident failure diversity). Four experiments have been designed to examine the relationship between the accuracy of the team and the measures of diversity, and among the measures themselves. Although there are proven connections between diversity and accuracy in some special cases, our results raise some doubts about the usefulness of diversity measures in building classifier ensembles in real-life pattern recognition problems.     

Decision fusion systems for fault detection and identification in industrial processes

Numerous fault detection and identification methods have been developed in recent years, whereas, each method works under its own assumption, which means a method works well in one condition may not provide a satisfactory performance in another condition. In this paper, we intend to design a fusion system by combining results of various methods. To increase the diversity among different methods, the resampling strategy is introduced as a data preprocessing step. A total of six conventionally used methods are selected for building the fusion system in this paper. Decisions generated from different models are combined together through the Dempster-Shafer evidence theory. Furthermore, to improve the computational efficiency and reliability of the fusion system, a new diversity measurement index named correlation coefficient is defined for model pruning in the fusion system. Fault detection and identification performances of the decision fusion system are evaluated through the Tennessee Eastman process.     

基于信任函数理论的修正融合目标识别算法

针对信任函数理论中经典Dempster组合规则难以有效融合高冲突证据并存在焦元基模糊问题,提出了一种基于信任函数理论的修正融合目标识别算法.修正融合算法在对相容命题进行组合时,考虑了焦元基的影响,使基本信任质量合理地向基数较小的焦元命题聚焦,以避免焦元基模糊问题;在对冲突命题进行组合时,对命题进行倾向性分析并对局部冲突采用局部分配的策略,以有效融合高冲突证据.算例与仿真比较分析验证了此修正融合目标识别算法的合理有效性和优越性.     

D-S证据理论在数据融合中失效问题的研究

D S方法作为一种重要的处理不确定性问题的数据融合方法,已经广泛应用于各种数据融合系统中。解决D S融合公式在处理冲突证据时的失效问题一直是研究的热点。国内外的各种改进方法主要分为对融合公式的改进和对融合模型的改进2个方向。对各种方法进行理论上和数据上的比较分析表明:修改模型的方式效果明显优于修改融合公式。     

基于信息融合的井下风险评估方法分析

随着近年来国家对煤矿安全的关注不断增强和煤矿信息化建设的发展,针对煤矿井下环境安全性评估和风险预测,提出了一种以信息融合理论为基础,利用Dempster—Shafer证据理论和粗糙集方法进行信息处理,解决煤矿对环境评估的不确定性和不精确性等问题,并综合考虑煤矿井下环境复杂性的特点,使用多传感器相关参数,实现了对井下多种信息的融合和算法的优化.该方法整体上提高了煤矿环境评估的有效性和监测数据的可靠性,并实现了对井下安全趋势的预测。     

多传感器多级数据融合的敌我识别估计

介绍了指挥、控制、通信和情报(C3I)系统多传感器数据融合的特点和一种空战融合模式,结合实例说明了多传感器各自对目标信息的测量结果进行融合处理,以期得到对目标属性的准确估计。     

基于DSP的航姿系统多传感器信息融合技术

设计了基于DSP的专用导航计算机,并以此为硬件平台,采集陀螺仪、加速度计、磁航向传感器和速度传感器信号,利用卡尔曼滤波技术进行多传感器信息融合,成功搭建了低成本小型航姿系统。针对该航姿系统的特点,设计了导航计算机程序快速更新软件,对卡尔曼滤波器进行低阶处理。针对导航计算机“数字信号处理器(DSP)+单片机(MCU)”的特殊结构,设计了合理的多传感器信息融合程序。实验证明:航姿系统利用多传感器信息融合技术,使用自行研制的专用导航计算机平台,姿态误差小于0.2,°航向误差小于0.5°,且大大减小了系统成本、体积和功率,具有实际应用价值。     

A kernel machine-based secure data sensing and fusion scheme in wireless sensor networks for the cyber-physical systems

Wireless sensor networks (WSNs) as one of the key technologies for delivering sensor-related data drive the progress of cyber-physical systems (CPSs) in bridging the gap between the cyber world and the physical world. It is thus desirable to explore how to utilize intelligence properly by developing the effective scheme in WSN to support data sensing and fusion of CPS. This paper intends to serve this purpose by proposing a prediction-based data sensing and fusion scheme to reduce the data transmission and maintain the required coverage level of sensors in WSN while guaranteeing the data confidentiality. The proposed scheme is called GM–KRLS, which is featured through the use of grey model (GM), kernel recursive least squares (KRLS), and Blowfish algorithm (BA). During the data sensing and fusion process, GM is responsible for initially predicting the data of next period with a small number of data items, while KRLS is used to make the initial predicted value approximate its true value with high accuracy. The KRLS as an improved kernel machine learning algorithm can adaptively adjust the coefficients with every input, while making the predicted value more close to actual value. And BA is used for data encoding and decoding during the transmission process due to its successful applications across a wide range of domains. Then, the proposed secure data sensing and fusion scheme GM–KRLS can provide high prediction accuracy, low communication, good scalability, and confidentiality. In order to verify the effectiveness and reasonableness of our proposed approach, we conduct simulations on actual data sets that are collected from sensors in the Intel Berkeley research lab. The simulation results have shown that the proposed scheme can significantly reduce redundant transmissions with high prediction accuracy.     

两种入侵检测系统D-S证据理论融合算法的比较

D-S证据理论可以支持概率推理、诊断、风险分析以及决策支持等,并在多传感器网络、医疗诊断等应用领域内得到了具体应用。该文比较了入侵检测中基于D-S证据理论的两种融合算法,通过DARPA1999评测数据集验证,得出各自的利弊,并提出了改进建议。     

基于模糊神经网络和证据理论的列车智能控制系统多信息融合故障诊断研究

针对列车智能控制系统故障诊断中的多故障特征信息输入时的时变、冗余、不确定性和空间分布性,给出了一种列车智能控制系统多信息融合故障诊断的系统结构。讨论了采用模糊神经网络进行特征层融合和证据理论进行决策层融合相结合的列车智能控制系统多信息融合故障诊断方法。故障诊断实例的结果表明:该方法能够有效地提高诊断的可信度,减小诊断的不确定性。     

基于支持向量机的多维度数据融合在油气预测中应用磁

油气预测中存在诸多不确定因素,包括属性的选取、预测算法的选取等。为了尽量消除这些不确定因素,论文利用支持向量机与信息融合理论进行地震油气预测,通过使用支持向量机内积函数定义的变换将输入空间映射到高维空间,并运用信息融合与支持向量机相结合的方法,尽可能的减小多种因素所带来的不确定性,提高油气预测精度,并通过实际数据验证,得到很好的效果。     

On the influence of tool geometry on robot motions

This paper describes a new algorithm for analysing the influence of the tool geometry on the motions of industrial robots in large manufacturing systems. On the basis of given points which are defined by the periphery devices used, the proposed method allows an evaluation of the cycle time. The computational time which is needed to determine the robot trajectory is kept small because an approach with constant velocities is calculated instead of the real path. In addition to an analysis of a single motion between two points, it is possible to calculate values for motions between several points which are linked together by a point to point move, so that an optimization of complex processes can be reached.     

基于AR的国家开放大学计算机专业体验式教学软件的设计开发

AR是一种新型多媒体计算机技术,融模式识别技术、机器学习技术、多媒体现实技术、3D建模技术于一体,创建一个具有智能性、交互性、沉浸感的虚实结合的系统。计算机专业体验式教学软件的设计对AR技术的应用与传播提供了有力支持,能够根据实际需求创设一个装配式开发环境,对当前所使用的软件开发环境进行扩展与延伸,借助多种类型建模组件将软件开发过程直观地展现出来,无需购置或者构建大型应用环境,将虚拟世界与真实世界无缝链接,从而提高国家开放大学学生的学习效率。