Description: 数据挖掘算法之一，基于代表点的CURE聚类算法，该算法先把每个数据点看成一类，然后合并距离最近的类，直至类个数为所要求的个数为止。-CURE cluster algorithm based on representive point,one of data mining algorithms,classifies each data as a category firstly, then unifies categories with the nearest distance into one until the number of class is coincidence with the classes demanded. Platform: |
Size: 45056 |
Author:黄镇 |
Hits:
Description: 模式识别的经典算法之一，动态聚类的k均值算法，采用matlab进行编程，并对分类进行了画图分析。-the classic pattern recognition algorithms, dynamic clustering algorithm k mean using Matlab programming, as well as classification of the class analysis. Platform: |
Size: 2048 |
Author:也风 |
Hits:
Description: 实现聚类K均值算法： K均值算法：给定类的个数K，将n个对象分到K个类中去，使得类内对象之间的相似性最大，而类之间的相似性最小。-achieving K-mean clustering algorithms : K-means algorithm : given the number of Class K, n objects assigned K to 000 category, making such objects within the similarity between the largest category of the similarity between the smallest. Platform: |
Size: 1024 |
Author:yili |
Hits:
Description: k-means聚类算法在二维平面上的可视化实现
聚类时可以设置类数和迭代阈值
聚类结果用色彩和类圆清楚的表现出来-k-means clustering algorithm in a two-dimensional plane with the Visualization of clustering can be set up several categories and iterative threshold Clustering results using color and class round clearly demonstrated Platform: |
Size: 45056 |
Author:周黎明 |
Hits:
Description: A general technique for the recovery of signi cant
image features is presented. The technique is based on
the mean shift algorithm, a simple nonparametric pro-
cedure for estimating density gradients. Drawbacks of
the current methods (including robust clustering) are
avoided. Feature space of any nature can be processed,
and as an example, color image segmentation is dis-
cussed. The segmentation is completely autonomous,
only its class is chosen by the user. Thus, the same
program can produce a high quality edge image, or pro-
vide, by extracting all the signi cant colors, a prepro-
cessor for content-based query systems. A 512 512
color image is analyzed in less than 10 seconds on a
standard workstation. Gray level images are handled
as color images having only the lightness coordinate-A general technique for the recovery of sig ni cannot image features is presented. The techni que is based on the mean shift algorithm, a simple nonparametric pro-cedure for estimat ing density gradients. Drawbacks of the curren t methods (including robust clustering) are av oided. Feature space of any nature can be proces sed, and as an example, color image segmentation is dis-cussed. The se gmentation is completely autonomous. only its class is chosen by the user. Thus, the same program can produce a high quality edge image, or pro-vide. by extracting all the signi cannot colors, a prepro- cessor for content-based query syste ms. A 512,512 color image is analyzed in less than 10 seconds on a standard workstation. Gray 4ISR l images are handled as color images having only the lightness c Platform: |
Size: 17408 |
Author:gggg |
Hits:
Description: 一个简单的分级聚类算法，数据结构采用了VC++中提供的CArray类，可以任意添加样本，删除和修改样本，还可以将聚类结果保存起来，对学习聚类算法和VC++中文件操作及控件有一定的参考意义-A simple hierarchical clustering algorithm, data structure used VC++ Provided CArray class can add samples, delete and modify the sample clustering results can also be saved, and the clustering algorithm for learning VC++ document operation and control of some reference significance Platform: |
Size: 162816 |
Author:崔林艳 |
Hits:
Description: state of art language modeling methods:
An Empirical Study of Smoothing Techniques for Language Modeling.pdf
BLEU, a Method for Automatic Evaluation of Machine Translation.pdf
Class-based n-gram models of natural language.pdf
Distributed Language Modeling for N-best List Re-ranking.pdf
Distributed Word Clustering for Large Scale Class-Based Language Modeling in.pdf
-state of art language modeling methods: An Empirical Study of Smoothing Techniques for Language Modeling.pdfBLEU, a Method for Automatic Evaluation of Machine Translation.pdfClass-based n-gram models of natural language.pdfDistributed Language Modeling for N-best List Re-ranking . pdfDistributed Word Clustering for Large Scale Class-Based Language Modeling in.pdf Platform: |
Size: 2016256 |
Author:wen6860 |
Hits:
Description: 统计模式识别工具箱（Statistical Pattern Recognition Toolbox）包含：
1，Analysis of linear discriminant function
2，Feature extraction: Linear Discriminant Analysis
3，Probability distribution estimation and clustering
4，Support Vector and other Kernel Machines-
This section should give the reader a quick overview of the methods implemented in
STPRtool.
• Analysis of linear discriminant function: Perceptron algorithm and multiclass
modification. Kozinec’s algorithm. Fisher Linear Discriminant. A collection
of known algorithms solving the Generalized Anderson’s Task.
• Feature extraction: Linear Discriminant Analysis. Principal Component Analysis
(PCA). Kernel PCA. Greedy Kernel PCA. Generalized Discriminant Analysis.
• Probability distribution estimation and clustering: Gaussian Mixture
Models. Expectation-Maximization algorithm. Minimax probability estimation.
K-means clustering.
• Support Vector and other Kernel Machines: Sequential Minimal Optimizer
(SMO). Matlab Optimization toolbox based algorithms. Interface to the
SVMlight software. Decomposition approaches to train the Multi-class SVM classifiers.
Multi-class BSVM formulation trained by Kozinec’s algorithm, Mitchell-
Demyanov-Molozenov algorithm Platform: |
Size: 4271104 |
Author:查日东 |
Hits:
Description: ISODATA算法是一种基于统计模式识别的，非常经典的非监督学习动态聚类算法，有较强的实用性。ISODATA算法不仅可以通过调整样本所属类别完成样本的聚类分析，而且可以自动地进行类别的“合并”和“分裂”，从而得到类数比较合理的聚类结果。-ISODATA algorithm is based on statistical pattern recognition, and very classic dynamic clustering of non-supervised learning algorithm has good practicability. ISODATA algorithm can not only adjust the class to complete samples of cluster analysis of samples, and can automatically type of " merger" and " split" in order to get a few more reasonable type of clustering results. Platform: |
Size: 190464 |
Author:justinchan |
Hits:
Description: This paper presents a clustering approach
which estimates the specific subspace and the intrinsic dimension of each class. Our approach
adapts the Gaussian mixture model framework to high-dimensional data and estimates
the parameters which best fit the data. We obtain a robust clustering method called High-
Dimensional Data Clustering (HDDC). We apply HDDC to locate objects in natural images
in a probabilistic framework. Experiments on a recently proposed database demonstrate the
effectiveness of our clustering method for category localization. Platform: |
Size: 193536 |
Author:tra ba huy |
Hits:
Description: CVAP includes 4 External validity indices, 14 Internal validity indices and 5 clustering algorithms (K-means, PAM, hierarchical clustering, SOM and etc.). It supports other clustering algorithms via loading a solution file with class labels, or by adding new codes. And similarity metrics of Euclidean distance and Pearson correlation coefficient are supported.-CVAP includes 4 External validity indices, 14 Internal validity indices and 5 clustering algorithms (K-means, PAM, hierarchical clustering, SOM and etc.). It supports other clustering algorithms via loading a solution file with class labels, or by adding new codes. And similarity metrics of Euclidean distance and Pearson correlation coefficient are supported. Platform: |
Size: 258048 |
Author:tra ba huy |
Hits:
Description: This package contains Matlab m-files for learning finite Gaussian mixtures from sample data and performing data classification with Mahalanobis distance or Bayesian classifiers. Each class in training set is learned individually with one of the three variations of the Expectation Maximization algorithm: the basic EM algorithm with covariance fixing, the Figueiredo-Jain clustering algorithm and the greedy EM algorithm.
The basic EM and FJ algorithms can handle complex valued data directly, the greedy EM algorithm cannot.
Platform: |
Size: 20480 |
Author:李 |
Hits:
Description: C++中的ISODATA算法示例，ISODATA算法是一种基于统计模式识别、经典的动态聚类算法，有较强的实用性。ISODATA算法不仅可以通过调整样本所属类别完成样本的聚类分析，而且可以自动地进行类别的“合并”和“分裂”，从而得到类数比较合理的聚类结果。
-C++ examples of ISODATA algorithm, ISODATA algorithm is based on statistical pattern recognition, a classic of the dynamic clustering algorithm, there are more practical. ISODATA algorithm can not only by adjusting the samples to complete the class cluster analysis of samples, and can automatically type of " merger" and " split" in order to get a few more reasonable type of clustering results. Platform: |
Size: 11264 |
Author:jiangjihai |
Hits:
Description: the classic pattern recognition algorithms, dynamic clustering algorithm k mean using Matlab programming, as well as classification of the class analysis Platform: |
Size: 2048 |
Author:jntu |
Hits:
Description: 珍藏版，可实现，新K均值聚类算法，分为如下几个步骤：
一、初始化聚类中心
1、根据具体问题，凭经验从样本集中选出C个比较合适的样本作为初始聚类中心。
2、用前C个样本作为初始聚类中心。
3、将全部样本随机地分成C类，计算每类的样本均值，将样本均值作为初始聚类中心。
二、初始聚类
1、按就近原则将样本归入各聚类中心所代表的类中。
2、取一样本，将其归入与其最近的聚类中心的那一类中，重新计算样本均值，更新聚类中心。然后取下一样本，重复操作，直至所有样本归入相应类中。
三、判断聚类是否合理
采用误差平方和准则函数判断聚类是否合理，不合理则修改分类。循环进行判断、修改直至达到算法终止条件。-NewK-means clustering algorithm ，Divided into the following several steps:
A, initialize clustering center
1, according to the specific problems, from samples with experience selected C a more appropriate focus the sample as the initial clustering center.
2, with former C a sample as the initial clustering center.
3, will all samples randomly divided into C, calculate the sample mean, each the sample mean as the initial clustering center.
Second, initial clustering
1, according to the sample into the nearest principle clustering center represents the class.
2, as this, take the its recent as clustering center of that category, recount the sample mean, update clustering center. And then taking off, as this, repeated operation until all samples into the corresponding class.
Three, judge clustering is reasonable
Adopt error squares principles function cluster analysis.after clustering whether reasonable, no reasonable criterion revisio Platform: |
Size: 1024 |
Author:姜亮 |
Hits:
Description: 非线性分类器，对于任意输入的2维向量采用聚类算法分类-Non-linear classifier, for any type of two-dimensional vector classification using clustering algorithms Platform: |
Size: 63488 |
Author:李伟 |
Hits:
Description: k 均值聚类算法 ，能有效的将数据分成k类
但是具有k参数难以确定的缺点。
-k-means algorithm can cluster data into K class but, the parameter K can not be selected easily. Platform: |
Size: 5120 |
Author:cluster |
Hits:
Description: 基于神经网络的遥感图像分类取得了较好的效果,但存在固有的过学习、易陷入局部极小等缺点.支持向量机机器学习方法,根据结构风险最小化(SRM)原理,表现出很多优于其他传统方法的性能,本研究的基于多类支持向量机分类器的遥感图像分类取得了达95.4 的分类精度.但由于遥感图像分类类别多,所需训练样本较大,人工选择效率较低,为此提出以人工选择初始聚类质心、C均值模糊聚类算法自动标注训练样本的基于多类支持向量机的半监督式遥感图像分类方法,期望能在获得适用的分类精度的基础上有效提高分类效率-Neural net based remote sensing image classification has obtained good results. But neural net has inherent
flaws such as overfitting and local minimums. Support vector machine (SVM), which is based on Structural Risk Min-
imization(SRM), has shown much better performance than most other existing machine learning methods. Using mul-
ti-class SVM classifier high class rate of 95.4 is obtained. But for the class number of remote sensing image is much
great, manually obtaining of training samples is a much time-consuming work. So a multi-class SVM based semi-super-
vised approach is presented. It is choosed that the initial clustering centroids manually first, then label the samples as
the training ones automatically with fuzzy clustering algorithm. It is believed that this method will upgrade the classifi-
cation efficiency greatly with practicable class rate Platform: |
Size: 25600 |
Author:cissy |
Hits:
Description: 完成了一个基本的蚁群聚类算法，根据蚁群活动，信息数残留，类进行聚类-Completed a basic ant colony clustering algorithm based on ant colony activities, information on the number of residues, class clustering Platform: |
Size: 78848 |
Author:Minghui Pan |
Hits:
Search - class clustering