特征向量

In view of the the difficulty of mixture gas's characteristic absorption spectrum line overlapped, a novel seriatim mixture gas recognition method based on Support Vector Machine two classes classification and recognition model was presented.

针对混合气体种类光谱识别中组分气体特征谱线重叠严重的问题，将支持向量机用于混合气体种类光谱识别中，提出了一种基于支持向量机二值分类识别模型的逐一混合气体种类识别方法。

Moreover SVM, a machine learning method is introduced to study the features of Chinese person names recognition. By the aid of this statistic classification technique, it is proved that SVM is also feasible to recognize Chinese person names.

此外，论文还引入了支持向量机学习方法，从统计学分类的角度探讨中文姓名与非中文姓名在若干特征上的差异，验证应用支持向量机理论识别中文姓名的可行性。

And finally,the dominant characteristics are extracted which are composed of representative vectors of faces through analyzing the eigencurves using Fourier describer.

作为对人面特征的描述，最后利用傅里叶描绘子对特征曲线进行解析提取关键特征，得到人面的表征向量。

At first, the AdaBoost is used to detect human faces in the video, then the detected faces are normalized and projected to the incremental eigenspace to obtain their vector representation.

首先通过AdaBoost检测视频序列中的人脸图像，将检测到的人脸做标准化处理后投影到增量特征人脸子空间中，得到人脸图像的向量表述；然后应用单类支持向量机进行训练和分类，根据分类的结果动态地调整前面得到的最优分类超平面，实现对电影视频中特定演员的检索功能。

A fuzzy membership model based on support vector data description is proposed to fuzzify all the training data. Then the model is introduced into least square support vector machines. Data samples in input space are mapped into a high dimensional feature space, and then the smallest enclosing hypersphere is found.

将模糊隶属度概念引入最小二乘支持向量机，提出一种基于支持向量数据域描述的模糊隶属度函数模型，将输入空间中的样本映射到一个高维的特征空间；然后根据其偏离数据域的程度赋予不同的隶属度。

Base vectors are selected for formulating the nonlinear mapping. An approximate explicit expression is formed by using the Nystrm method for nonlinear mapping which is not available originally. The data in the nonlinear input space is mapped into some linear feature subspace and fisher linear discriminant analysis is computed for the mapped data.

首先提取基向量；然后采用Nystrm方法，以基向量为训练样本，将形式未知的非线性映射近似表达为已知形式的非线性映射，这种近似的非线性映射将变量由非线性的输入空间转换到线性的特征子空间；最后对映射数据进行线性Fisher判别分析。

Moreover, their intrinsic connection, advantages and disadvantages were analyzed from the view of multilinear projections, essential differences and their own superiorities between tensor and vector patterns were clarified, and the problems to be solved and future research were discussed.

对现有张量型学习算法进行总结：针对向量模式表示存在的问题，讨论了设计张量型学习算法的必要性和重要意义；从特征提取和分类器设计2个层面，对现有张量型学习算法进行分类和汇总，并着重从多线性投影方式出发，讨论了其内在联系及优缺点；分析了张量模式与向量模式表示间的本质关系及各自优势；探讨了张量型学习算法有待解决的问题及可能的研究趋势。

Since the size of image covariance matrices is same as that of images and the within-class image covariance matrix is usual nonsingular, thus, the difficulty resulting from high dimensionality and singular case are artfully avoided.

该文提出了一种基于复向量进行特征表不的并行融合新策略，建立了复线性投影分析的理论，给出了复特征空间内进行线性特征抽取方法，事实上，经典的线性投影分析理论与方法只是本文理论与方法的一个特例。

Also how to calculate normal vector of the original surface is introduced.

在第三章，介绍了特征的准确定义，并介绍了特征提取算法，并在特征提取的算法中，介绍了一种计算网格曲面顶点法向量的比较好的方法。

First, used the template examination law to pick up character point of specific character information got from the pretreatment of fingerprint picture; then, used using the crestline track technology to filter the false characteristic point; at last, used the improved partial vector pattern matching method to match fingerprint, and made a judge of whether it belonged to the same person.

首先是对指纹图像预处理得到的细节特征信息用模板检测法来提取特征点，然后用脊线跟踪技术滤除伪特征点，最后采用改进的局部向量点模式匹配法来对指纹进行匹配，判断是否同属一人。

In the negative and interrogative forms, of course, this is identical to the non-emphatic forms.

。但是，在否定句或疑问句里，这种带有"do"的方法表达的效果却没有什么强调的意思。

Go down on one's knees;kneel down

屈膝跪下。。。下跪祈祷