满意特征选择及其应用

实际应用中的特征选择是一个满意优化问题.针对已有特征选择方法较少考虑特征获取代价和特征集维数的自动确定问题,提出一种满意特征选择方法(SFSM),将样本分类性能、特征集维数和特征提取复杂性等多种因素综合考虑.给出特征满意度和特征集满意度定义,设计出满意度函数,导出满意特征集评价准则,详细描述了特征选择算法.雷达辐射源信号特征选择与识别的实验结果显示,SFSM在计算效率和选出特征的质量方面明显优于顺序前进法、新特征选择法和多目标遗传算法.证实了SFSM的有效性和实用性.     

基于全融合网络的三维点云语义分割

为提高室内场景的点云语义分割精度,设计了一个全融合点云语义分割网络。网络由特征编码模块、渐进式特征解码模块、多尺度特征解码模块、特征融合模块和语义分割头部组成。特征编码模块采用逆密度加权卷积作为特征编码器对点云数据进行逐级特征编码,提取点云数据的多尺度特征;然后通过渐进式特征解码器对高层语义特征进行逐层解码,得到点云的渐进式解码特征。同时,多尺度特征解码器对提取的点云多尺度特征分别进行特征解码,得到点云多尺度解码特征。最后将渐进式解码特征与多尺度解码特征融合,输入语义分割头部实现点云的语义分割。全融合网络增强了网络特征提取能力的鲁棒性,实验结果也验证了该网络的有效性。     

基于正余弦算法的文本特征选择

为获取文本中的较优特征子集,剔除干扰和冗余特征,提出了一种结合过滤式算法和群智能算法的混合特征寻优算法。首先计算每个特征词的信息增益值,选取较优的特征作为预选特征集合,再利用正余弦算法对预选特征进行寻优,获取精选特征集合。为较好地平衡正余弦算法中的全局搜索和局部开发能力,加入了自适应惯性权重;为更精确地评价特征子集,引入以特征数量和准确率进行加权的适应度函数,并提出了新的位置更新机制。在KNN和贝叶斯分类器上的实验结果表明,该特征选择算法与其它特征选择算法及改进前的算法相比,分类准确率得到了一定的提升。     

基于局部特征识别的特征有效性维护方法

缺乏特征模型的有效性维护功能已经成为目前特征造型系统存在的一个严重而亟待解决的问题.在对特征有效性条件进行深入分析的基础上,提出了一个基于扩展属性邻接图（extended attributed adjacency graph,简称EAAG）的特征有效性表示方法,特别是提出了基于局部特征识别的特征有效性维护新方法.该方法不仅能够自动判别特征的有效性是否被破坏,而且能确定导致特征无效的原因和遭破坏后特征的状态,从而能够根据用户的意图自动维持特征模型的有效性.     

基于特征面的分层特征识别

为降低特征识别的复杂度,提出基于特征实体、特征实面和特征虚面概念的层次性特征分类方法.通过构造2类神经网络输入矩阵,利用神经网络在特征识别中所具有的优势,实现基于特征面的分层特征识别方法.实例表明：该方法在识别去除材料的特征时比较有效,但识别特征的范围受到一定限制.     

一种新的隐马尔可夫模型及其在手绘图形识别中的应用

提出了一种新的隐马尔可夫模型——自适应隐马尔可夫模型（AHMM）．与传统的开环HMM相区别,AHMM是一种用于识别的带反馈机制的闭环HMM．AHMM采用带有压缩率调整因子的特征压缩算法,首先对待识别的特征序列进行较高压缩率的压缩,然后将压缩得到的特征序列送入HMM识别器进行识别．根据对识别效果满意度的判决,确定是否需要调整压缩率因子以获得较长的特征序列,并重新送入HMM识别器进行识别．将该文提出的AHMM用于联机手绘图形的识别,实验表明,AHMM方法与传统的HMM方法相比,识别率和识别速度均有显著提高．     

Ultrasonic liver tissue characterization by feature fusion

This paper describes a two-stage feature fusion method for ultrasonic liver tissue characterization. The proposed method hierarchically incorporates a genetic-algorithm-based feature selection to automatically select more efficient feature subset to discriminate among ultrasonic images of liver tissue in three states: normal liver, cirrhosis, and hepatoma. Multiple feature spaces are adopted in this paper, including the spatial gray-level dependence matrices (SGLDMs), multiresolution fractal feature vector and multiresolution energy feature vector. Features extracted from different feature spaces may contain complementary information. The feature subsets of different feature spaces are fused and the genetic-algorithm-based feature selection is applied onto the fused feature space to facilitate the two-stage feature fusion. The classification accuracy of the fused feature subset is up to 96.62%. Experimental results demonstrate that the proposed method is capable to select discriminative features among multiple feature vectors to achieve the early detection of hepatoma and cirrhosis based on ultrasonic liver imaging.     

自由曲面特征的定义与识别技术的研究

结合自主开发的HUST-CAID（哈尔滨理工大学计算机辅助工业造型设计）系统的特点,引入自由曲面特征,给出了自由曲面特征的参数化定义,并在此基础上对自由曲面特征识别进行研究,先是给出自由曲面特征识别的定义,接着提出了基于曲线的特征识别的算法。该方法将基于曲线的特征以二维参数的形式给出定义,使其能在二维平面上研究,通过重构特征的剖面模板库,从而利用目标特征与模板特征匹配实现了特征识别。     

特征采样和特征融合的子图像人脸识别方法

提出一种基于特征采样和特征融合的子图像人脸识别方法(RS-SpCCA).首先,对子图像进行特征采样;然后,将全局特征和采样后的特征使用CCA进行信息融合,以获取包含全局特征和局部特征的相关特征;最后,在相关特征上构建分量分类器.在该方法中,特征采样是为了构建更多且多样的分量分类器;而引入特征融合思想是为了充分利用图像的全局特征.AR,Yale和ORL这3个数据库上的实验结果表明,基于特征采样和特征融合的子图像方法(RS-SpCCA)优于单纯的信息融合方法(SpCCA)和特征采样方法(Semi-RS).     

面向三维变量设计的可变特征模型

本文提出了一个面向三维变量设计的产品形状可变特征模型，其中包括特征树，特征表有特征约束关系图等新概念，用于描述设计过程中的形体模型和特征间的相互内在联系，从而不仅可作参数化设计，还支持更广泛意义上的变量设计。     

Maintaining associativity between form feature models

The promise of features technology was that the task domains would have access to task specific product data through feature based models. This is an important requirement in a distributed and concurrent design environment, where data of part geometry has to be shared between different task domains.Associativity between feature models implies the automatic updating of different feature models of a part after changes are made in one of its feature models. The proposed algorithm takes multiple feature models of a part as input and modifies other feature models to reflect the changes made to a feature in a feature model. The proposed algorithm updates feature volumes in other feature models and then classifies the updated volumes to obtain the updated feature model. The spatial arrangement of feature faces and adjacency relationship between features are used to isolate features in a view that are affected by the modification. Feature volumes are updated based on the classification of the feature volume of the modified feature with respect to feature volumes of the model being updated. The algorithm is capable of handling all types of feature modifications namely, feature deletion, feature creation, and changes to feature location and parameters. In contrast to current art in automatic updating of feature models, the proposed algorithm does not use an intermediate representation, does not re-interpret the feature model from a low level representation and handles interacting features. The present work considers modifications to form features only. Modification of constraints and application attributes are under investigation. Results of implementation on typical cases are presented.     

一种组合特征抽取的新方法

该文提出了一种基于特征级融合的特征抽取新方法，首先，给出了一种合理的特征融合策略，即利用复向量给出组合特征的表示，将特征空间从实向量空间拓广到复向量空间，然后，发展了具有统计不相关性的鉴别分析的理论，并将其用于复向量空间内最优鉴别特征的抽取，最后，在Concordia大学的CENPARMI手写体阿拉伯数字数据库以及南京理工大学NUST603HW手写汉字库上的试验结果表明，所提出的组合特征抽取方法不仅具有很强的维数压缩能力，而且较大幅度地提高了识别率。     

基于动静态组合特征参数的语音识别

基于语音信号的时变特性,本文提出了动静态特征参数结合的语音信号识别方法,首先在特征参数提取中引入了小波包变换,借助MFCC(Mel-Frequency Cepstrum Coefficient)参数的提取方法,用小波包变换代替傅立叶变换和Mel滤波器组,提取了新的静态特征参数DWPTMFCC(Discrete Wavelet Packet Transform Mel-Frequency Coefficient),然后把它与一阶DWPTMFCC差分参数相结合成一个向量,作为一帧语音信号的参数,通过试验和仿真,此参数具有很高的识别率,是一种很好的语音特征参数.并且把混沌特性引入到神经元,构成混沌神经网络,把这种神经网络用于语音识别,并与常用的BP神经网络识别方法进行了比较.试验结果表明,混沌神经网络的平均识别率要高于同等条件下常用的神经网络方法的识别率.     

An improved global feature selection scheme for text classification

Feature selection is known as a good solution to the high dimensionality of the feature space and mostly preferred feature selection methods for text classification are filter-based ones. In a common filter-based feature selection scheme, unique scores are assigned to features depending on their discriminative power and these features are sorted in descending order according to the scores. Then, the last step is to add top-N features to the feature set where N is generally an empirically determined number. In this paper, an improved global feature selection scheme (IGFSS) where the last step in a common feature selection scheme is modified in order to obtain a more representative feature set is proposed. Although feature set constructed by a common feature selection scheme successfully represents some of the classes, a number of classes may not be even represented. Consequently, IGFSS aims to improve the classification performance of global feature selection methods by creating a feature set representing all classes almost equally. For this purpose, a local feature selection method is used in IGFSS to label features according to their discriminative power on classes and these labels are used while producing the feature sets. Experimental results on well-known benchmark datasets with various classifiers indicate that IGFSS improves the performance of classification in terms of two widely-known metrics namely Micro-F1 and Macro-F1.     

融合特征增强和自注意力的SSD小目标检测算法

SSD是一种多尺度目标检测算法,由于浅层特征图缺乏语义信息,导致小目标的检测准确率低.针对这个问题,提出一种融合特征增强和自注意力的SSD小目标检测算法FA-SSD.该算法在SSD基础上构建一条自深向浅的递归反向路径,此路径包含三个模块:深层特征增强模块利用路径深层多尺度特征图生成的上下文信息和最深层特征图的语义信息,...     

参数化特征造型中拓扑结构变异的一种解决方法

文中首先分析了基于边界表示的特征模型用于参数化特征变动设计所存在的问题,指出了以边界表示中的几何边为形状特征定位约束基准无法满足拓拟结构变异的局限性,分析了基于草图的特征造型方法在变动设计中的重要性。在此基础上,提出了将基于实体模型边界表示的特征模型与基于草图的特征模型结合起来的新思想,并给出了该思想的实现方法,该方法使特征定位基准在特征造型过程中保持不变,且不随特征定位约束参数的改变而变化,能很好地满足参数化变动设计与拓扑结构变异的要求。     

Extracting feature lines from point clouds based on smooth shrink and iterative thinning

An index of measuring the variation on a surface called the smooth shrink index (SSI) which presents robustness to noise and non-uniform sampling is developed in this work. Afterwards, a new algorithm used for extracting the feature lines was proposed. Firstly, the points with an absolute value of SSI greater than a given threshold are selected as potential feature points. Then, the SSI is applied as the growth condition to conduct region segmentation of the potential feature points. Finally, a bilateral filter algorithm is employed to obtain the final feature points by thinning the potential feature points iteratively. While thinning the potential feature points, the tendency of the feature lines is acquired using principle component analysis (PCA) to restrict the drift direction of the potential feature points, so as to prevent the shrink in the endpoints of the feature lines and breaking of the feature lines induced by non-uniform sampling.     

Quantum inspired method of feature fusion based on von Neumann entropy

One mission of feature fusion is to obtain a complete yet concise presentation of all existing feature data by detecting and fusing the duplicate feature data. In contrast to the already developed feature fusion methods which have shown their limitations, this paper applies the theories of quantum information to feature fusion. Further, a novel and effective step-wise quantum inspired feature fusion method, which detects the duplicate feature data based on maximum von Neumann mutual information and fuses the duplicate feature data using the operations on quantum state, is developed. This same idea is also used for feature dimensionality reduction, and the corresponding models are investigated. For comparison, another quantum inspired feature fusion method based on average quantum phase is presented here. The experimental results show that the quantum inspired feature fusion method based on von Neumann entropy gives better results on completeness and conciseness than the method based on average quantum phase.     

基于特征集中度和差别对象对集的特征选择方法

本文首先简单分析了几种经典的特征选择方法,总结了它们的不足,然后提出了特征集中度的概念, 紧接着把差别对象对集引入粗糙集并提出了一个基于差别对象对集的属性约简算法,最后把该属性约简算法同特征 集中度结合起来,提出了一个综合性特征选择方法．该综合性方法首先利用特征集中度进行特征初选以过滤掉一些 词条来降低特征空间的稀疏性,然后再使用所提属性约简算法消除冗余,从而获得较具代表性的特征子集．实验结 果表明该综合性方法效果良好．     

Unsupervised feature selection using clustering ensembles and population based incremental learning algorithm

This paper describes a novel feature selection algorithm for unsupervised clustering, that combines the clustering ensembles method and the population based incremental learning algorithm. The main idea of the proposed unsupervised feature selection algorithm is to search for a subset of all features such that the clustering algorithm trained on this feature subset can achieve the most similar clustering solution to the one obtained by an ensemble learning algorithm. In particular, a clustering solution is firstly achieved by a clustering ensembles method, then the population based incremental learning algorithm is adopted to find the feature subset that best fits the obtained clustering solution. One advantage of the proposed unsupervised feature selection algorithm is that it is dimensionality-unbiased. In addition, the proposed unsupervised feature selection algorithm leverages the consensus across multiple clustering solutions. Experimental results on several real data sets demonstrate that the proposed unsupervised feature selection algorithm is often able to obtain a better feature subset when compared with other existing unsupervised feature selection algorithms.