基于集合经验模态分解和随机森林的短时交通流预测

针对短时交通流数据具有非线性、不确定性等特点,提出一种基于集合经验模态分解(ensemble empirical mode decomposition,EEMD)和随机森林(random forest,RF)的组合预测模型。首先,利用EEMD算法将原始交通流数据的区间平均速度序列分解为若干个本征模函数（intrinsic mode function,IMF）和一个残差分量（residual,RES),提取出交通流数据在不同时频的信息;接着,将第一个分量进行二次EEMD分解,细化交通流的随机信息;然后,将分解得到的各个分量分别使用RF进行预测,构建子模型;最后,将所有子模型的预测值线性求和,得到最终的预测结果。采用阿拉尔市某路段的实际交通流数据进行实验,结果表明,EEMD和RF的组合预测模型优于单一的RF模型,并且对IMF1进行二次EEMD分解可进一步提高组合预测模型的准确率。     

集成学习的多分类器动态融合方法研究

AdaBoost集成学习方法中,分类器一经学习成功,其投票权值就已确定,同一分类器对所有待测样本均有相同的投票权值。对于难于分类样本,具有良好分类性能的少数分类器权值却较低。提出适用于集成学习方法的权重自适应调整多分类器集成算法。根据多分类器行为信息,产生待测样本局部分类精度的有效判定区域,基于有效判定区域选择不同的分类器组合,并调整其相应权重,利用样本集上的统计信息来动态指导分类集成判决。实验结果表明,该算法提高了集成分类性能。     

Support Vector Machine Ensemble Based on Genetic Algorithm

Support vector machines （SVMs） have been introduced as effective methods for solving classification problems. However, due to some limitations in practical applications, their generalization performance is sometimes far from the expected level. Therefore, it is meaningful to study SVM ensemble learning. In this paper, a novel genetic algorithm based ensemble learning method, namely Direct Genetic Ensemble （DGE）, is proposed. DGE adopts the predictive accuracy of ensemble as the fitness function and searches a good ensemble from the ensemble space. In essence, DGE is also a selective ensemble learning method because the base classifiers of the ensemble are selected according to the solution of genetic algorithm. In comparison with other ensemble learning methods, DGE works on a higher level and is more direct. Different strategies of constructing diverse base classifiers can be utilized in DGE. Experimental results show that SVM ensembles constructed by DGE can achieve better performance than single SVMs, hagged and boosted SVM ensembles. In addition, some valuable conclusions are obtained.     

Moisture Absorption, Mass Loss and Evaporative Resistance of Clothing in Transient Condition

The water vapour resistance of clothing ensembles is not as commonly determined as dry thermal insulation. The measurement techniques are more complicated and the measurement values differ among laboratories. Due to complicated moisture transfer process through clothing ensemble, the moisture absorbed and evaporated varies in transient and steady state phases depending on properties, thickness of clothing, and environmental conditions. The purpose of this study was to measure moisture gain inside hygroscopic underwear, and hydrophobic and permeable outer wear as a function of time, to investigate mass loss from ＂skin＂ as well as from the manikin, to quantify evaporative heat loss and total heat loss from the manikin, and to determine water vaponr resistance of clothing. Manikin Tore was used by wearing wet ＂skin＂ to simulate sweating condition. Moisture content in the inner garment gain shows an exponential relation against time. Moisture in the outer permeable layer shows little gain. On the contrary, mass loss directly from the wet skin decreases exponentially. The mass loss from the manikin is relatively stable throughout three test phases. The evaporative heat loss is about 2/3 of the total heat loss from the sweating manikin. While measuring the evaporative resistance of clothing ensembles with hygroscopic inner garment and permeable outer garment, unlike to measure ensembles with impermeable outer layer, one hour measurement time is enough to get relatively stable results. The variation between the 1^rt hour and the 3^nd hour is less than 5%. The length of transient period and measurement time is dependent on the permeability, thickness of clothing ensembles and environmental conditions.     

一种基于级联模型的类别不平衡数据分类方法

真实世界问题中,不同类别的样本在数目上往往差别很大,而传统机器学习方法难以对小类样本进行正确分类,若小类的样本是足够重要的,就会带来较大的损失.因此,对类别分布不平衡数据的学习已成为机器学习目前面临的一个挑战.受计算机视觉中级联模型的启发,提出一种针对不平衡数据的分类方法BalanceCascade.该方法逐步缩小大类别使数据集趋于平衡,在此过程中训练得到的一系列分类器通过集成方式对预测样本进行分类.实验结果表明,该方法可以有效地提高在不平衡数据上的分类性能,尤其是在分类性能受数据的不平衡性严重影响的情况下.     

Learning to Predict in Complex Biological Domains

1 Introduction1 Bioinformatics is a rapidly expanding field involving a significant contribution from the data mining community. Many different problems are grouped into the general category of bioinformatics, including protein secondary structure prediction, drug discovery, DNA microarray analysis, gene prediction and genome analysis [17]. In this paper, we focus on two particular problems in which data mining can contribute: drug discovery and protein secondary structure prediction. Both …     

吉布斯佯谬的正则系综讨论

本文就近独立的离域子 (理想气体 )体系的混合熵所产生的吉布斯佯谬现象 ,运用正则系综的统计方法进行了简单分析和讨论     

A Robust Pedestrian Detection Approach Based on Shapelet Feature and Haar Detector Ensembles

Detection of pedestrians in images and video sequences is important for many applications but is very challenging due to the various silhouettes of pedestrians and partial occlusions. This paper describes a two-stage robust pedestrian detection approach. The first stage uses a full body detector applied to a single image to generate pedestrian candidates. In the second stage, each pedestrian candidate is verified with a detector ensemble consisting of part detectors. The full body detector is trained based on improved shapelet features, while the part detectors make use of Haar-like wavelets as features. All the detectors are trained by a boosting method. The responses of the part detectors are then combined using a detector ensemble. The verification process is formulated as a combinatorial optimization problem with a genetic algorithm for optimization. Then, the detection results are regarded as equivalent classes so that multiple detections of the same pedestrian are quickly merged together. Tests show that this approach has a detection rate of over 95% for 0.1% FPPW on the INRIA dataset, which is significantly better than that of the original shapelet feature based approach and the existing detector ensemble approach. This approach can robustly detect pedestrians in different situations.     

长程相互作用Blume-Emery-Griffiths模型:微正则系综的临界指数

我们研究了长程Blume-Emery-Griffiths模型在微正则系综下的热力学量,发现在一级相变附近不仅出现负比热,磁化率也呈现负值,并且温度、比热和磁化率在一级相变点都不连续.当系统发生有序-无序的二级相变时,比热的临界指数a=a′=0和磁化率的临界指数γ=γ′=1与平均场的结果一致.当耦合常数对应于正则系综的三临界点时,比热的临界指数0=a≠a′=1和磁化率的临界指数1=γ≠γ′=2被获得.我们发现临界指数关系a′+β+γ′>2破坏Widom标度律和等价性.     

遥感影像多分类器集成的关键技术与系统实现

 为克服单一分类器在遥感影像分类精度和效率方面的限制,有必要构建多分类器系统,集不同分类器的优点,获得比单一分类器更高的精度。针对遥感影像的特点和分类的需求,在遥感影像多分类器集成系统需求分析和系统设计的基础上,运用IDL语言在ENVI遥感影像处理平台下实现系统开发。遥感影像多分类器集成系统的主要功能包括遥感影像文件处理、特征选择与提取、分类预处理、分类、多种模式的多分类器集成(固定组合模式、用户自定义模式、向导模式和推荐模式)等。通过分类实例对系统应用进行介绍,表明本系统能够有效地提高遥感影像分类精度。