Location:
Search - K-mean
Search list
Description: 改进的K-Means算法，通过改进传统K-Means算法，剔除远离中心均值的离散点，加快算法的收敛速度。-Improved K-Means algorithm, by improving the traditional K-Means algorithm, removing the mean of discrete points away from the center to accelerate the convergence speed.
Platform: |
Size: 1196032 |
Author: swim |
Hits:
Description: nano- X是一个著名的开放式源码嵌入式GUI 软件，目的是把图形视窗环境引入到运行Linux 的小型设备和平台上。nano-X使用了分层设计的思想，可移植性非常好，nano-X 的图形引擎能够运行在任何支持readpixel，writepixel，drawhorzline, drawvertline 和setpalette 的系统之上，在底层函数的支持之下，nano-X 支持新的Linux内核帧缓存结构，并基于framebuffer来实现图像的绘制。nanox应用程序运行时占用的资源较少，server只有100多K，精巧的设计并不代表功能的简陋，目前提供每像素1、2、4、8、16、24和32位的支持，另外还支持彩色显示和灰度显示. 目前支持的图形文件包括WINDOWS的GIF,JPEG,BMP、PNG、XPM和PBM、PGM、XPM格式。
本程序是MIPS平台下LINUX 2.6下使用NANO-X API 实现数码相框的源程序。-nano-X is a well-known open-source embedded GUI software, the purpose is to introduce graphical Windows environment to run Linux in the small devices and platforms. nano-X uses a hierarchical design ideas, portability is very good, nano-X s graphics engine can run on any supported readpixel, writepixel, drawhorzline, drawvertline and setpalette system above the support function at the bottom, nano-X supports the new Linux kernel frame buffer structure, and to achieve the image-based rendering framebuffer. nanox application run-time occupation of fewer resources, server is only 100 K, compact design does not mean that features simple, currently available 1,2,4,8,16,24 and 32-bit per pixel support, in addition to color display and grayscale support. Currently supported graphic file, including WINDOWS the GIF, JPEG, BMP, PNG, XPM, and PBM, PGM, XPM formats.
This program is under the LINUX 2.6 MIPS platform to use NANO-X API to achieve digital photo frame of the source.
Platform: |
Size: 1377280 |
Author: 何庭光 |
Hits:
Description: Performs Procustes point alignment on a group of point sets. Method rigidly aligns, shifts, and scales points to reduce mean square error.
Method is described in:
B. Klare, P Mallapragada, A.K. Jain, and K. Davis, "Clustering Face Carvings: Exploring the Devatas of Angkor Wat", in Proceedings International Conference on Pattern Recognition (ICPR), 2010.
http://www.cse.msu.edu/~klarebre/docs/ICPR_AW.pdf-Performs Procustes point alignment on a group of point sets. Method rigidly aligns, shifts, and scales points to reduce mean square error.
Method is described in:
B. Klare, P Mallapragada, A.K. Jain, and K. Davis, "Clustering Face Carvings: Exploring the Devatas of Angkor Wat", in Proceedings International Conference on Pattern Recognition (ICPR), 2010.
http://www.cse.msu.edu/~klarebre/docs/ICPR_AW.pdf
Platform: |
Size: 1024 |
Author: B |
Hits:
Description: 首先从n个数据对象任意选择 k 个对象作为初始聚类中心；而对于所剩下其它对象，则根据它们与这些聚类中心的相似度（距离），分别将它们分配给与其最相似的（聚类中心所代表的）聚类；然后再计算每个所获新聚类的聚类中心（该聚类中所有对象的均值）；不断重复这一过程直到标准测度函数开始收敛为止。一般都采用均方差作为标准测度函数. k个聚类具有以下特点：各聚类本身尽可能的紧凑，而各聚类之间尽可能的分开。-First, a data object from the n choose k objects as initial cluster centers and for the rest of the other objects, according to their similarity with the cluster center (distance), respectively, assign them to their most similar (represented by cluster center) clustering then calculated for each cluster center received a new clustering (all objects in the cluster mean) repeats this process until the convergence criteria begin until the measure function. Standard deviation is generally used as a standard measure function. K a cluster has the following characteristics: the cluster itself as a compact, but separated as much as possible between each cluster.
Platform: |
Size: 1024 |
Author: lx |
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: 实现图像的旋转、平移、缩放功能，实现K均值分类功能-Image rotation, pan, zoom function, the mean classification functions to achieve K
Platform: |
Size: 3855360 |
Author: yinyaqiu |
Hits:
Description: k-means 算法接受参数 k ；然后将事先输入的n个数据对象划分为 k个聚类以便使得所获得的聚类满足：同一聚类中的对象相似度较高；而不同聚类中的对象相似度较小。聚类相似度是利用各聚类中对象的均值所获得一个“中心对象”（引力中心）来进行计算的。-K-means algorithm accept parameter k Then will the n of prior input data object is divided into k clustering to make won clustering meet: the same clustering the object in the similarity is higher And different clustering the object in the smaller similarity. Clustering similarity is using the cluster of the object in the mean gained a "center object" (the center of gravity) to calculate.
Platform: |
Size: 1024 |
Author: 彭立军 |
Hits:
Description: 使用opencv使得对输入的影像每K张做平均，并且这k张影像必须是以每一帧开始的image往后的K张做平均，然后得到连续的经过平均的video-Opencv makes use of the input image do mean every K Zhang, and these k images must be the beginning of each frame image is back to do the K Zhang average, and then get through the average of a continuous video
Platform: |
Size: 20893696 |
Author: tianxianxian |
Hits:
Description: PAM（Partitioning Around Medoid,围绕中心点的划分）算法是是划分算法中一种很重要的算法，有时也称为k-中心点算法，是指用中心点来代表一个簇。PAM算法最早由Kaufman和Rousseevw提出，Medoid的意思就是位于中心位置的对象。PAM算法的目的是对n个数据对象给出k个划分。PAM算法的基本思想：PAM算法的目的是对成员集合D中的N个数据对象给出k个划分，形成k个簇，在每个簇中随机选取1个成员设置为中心点，然后在每一步中，对输入数据集中目前还不是中心点的成员根据其与中心点的相异度或者距离进行逐个比较，看是否可能成为中心点。用簇中的非中心点到簇的中心点的所有距离之和来度量聚类效果，其中成员总是被分配到离自身最近的簇中，以此来提高聚类的质量。-PAM (Partitioning Around Medoid Around the division of the center,) algorithm is a kind of partition algorithm is very important algorithm, and sometimes also called k-center algorithm, it is to point to in the center to represent a cluster. The earliest PAM algorithm by Kaufman and Rousseevw puts forward, Medoid mean is at the center of the location of the object. PAM algorithm for the purpose of n data object is given k division. PAM algorithm to the basic idea of the: PAM algorithm for the purpose of members set D is the N data object given k division, forming k cluster, each cluster in selected at random from a members set to center, then at each step, the focus of the input data is not a member of the center according to the center YiDu or phase with each distance is, look to whether can be centered. Use cluster in the center point to the center of the cluster of the sum of all the distance to measure the clustering effect, which is always assigned members from their recent cluste
Platform: |
Size: 2048 |
Author: 赵元 |
Hits:
Description: 基于遗传算法的投影寻踪代码，提供大家下载，方便查阅-【研学堂】【代码】投影寻踪代码，请验用！！
function Qa=Project_Pursuit(X,a,Alpha)
输入参数列表
X 本指标矩阵，n×p的矩阵，每一行为一个样本，
Xij表示第i行第j列指标，X是否已经归一化均可
a 投影向量，1×p的矩阵，元素取值范围-1～1，要求其元素平方和等于1
Alpha 窗口半径系数，典型取值0.1
输出参数列表
Qa 投影指标函数
第零步：对a的预处理
b=sqrt(sum(a.^2))
a=a/b
第一步：归一化处理
[n,p]=size(X)
x=zeros(n,p)
Xjmax=max(X)
Xjmin=min(X)
for i=1:n
x(i,:)=(X(i,:)-Xjmin)./(Xjmax-Xjmin)
end
第二步：构造投影指标值
Z=zeros(n,1)
for i=1:n
Z(i)=sum(a.*x(i,:))
end
第三步：计算投影指标函数
计算类间类内矩阵散度
meanZ=mean(Z)
Sa=0
for k=1:n
sa=(Z(i)-meanZ(i)).^2
Sa=Sa+sa
Sa=sqrt(Sa/n)
end
R=Alpha*Sa 窗口半径
Da=0
for k=1:n
rik=abs(Z(i)-Z(k))
if R>rik
Da=Da+rik Da=Da+R-rik
end
end
Qa=Sa*Da
Platform: |
Size: 1024 |
Author: 余文乐 |
Hits:
Description: 数字信号处理的应用之一是从含有加性噪声的信号中去除噪声。现有被噪声污染的信号x[k]=s[k]+d[k]，式中： 为原始信号d[k]为均匀分布的白噪声。（1）分别产生50点的序列s[k]与白噪声序列ddd[k]，将二者叠加生成x[k]，并在同一张图上绘出x0[k]，d[k]与x[k]的序列波形。（2）均值滤波能有效去除叠加在低频信号上的噪声。已知3点滑动平均数字滤波器的单位脉冲响应为h[k]=[1，
-One of the applications of digital signal processing to remove noise from the signal with additive noise. Existing noise pollution signal x [k] = s [k] the+d [k], where: the original signal d [k] uniformly distributed white noise. (1), respectively, to produce a 50-point sequence s [k] white noise sequences ddd [k], the two superimposed to generate x [k], and plotted on the same graph x0 [k], d [k] and x [k] sequence waveform. (2) the mean filter can effectively remove noise superimposed on the low-frequency signals. Known 3-point moving average digital filter unit impulse response is h [k] = [1,
Platform: |
Size: 28672 |
Author: 推翻 |
Hits:
Description: FUNCTIONS :
STATISTICS
- MODA Mode of a distribution
- ARGMAX calculates position of the maximum value of matrix V
- ARGMIN calculates position of the minimum value of matrix V
- DRAW_HIST estimates and draws histogram of data
- Random k data generator
2D LOCAL OPERATORS
- LOCALMEAN local mean of 2D image
- VARLOCAL local variance of 2D image
- CVLOCAL Local Square Coefficient of variation
- LOCALMAD local Median Absolute Deviation (MAD)
- FUNCTIONS :
STATISTICS
- MODA Mode of a distribution
- ARGMAX calculates position of the maximum value of matrix V
- ARGMIN calculates position of the minimum value of matrix V
- DRAW_HIST estimates and draws histogram of data
- Random k data generator
2D LOCAL OPERATORS
- LOCALMEAN local mean of 2D image
- VARLOCAL local variance of 2D image
- CVLOCAL Local Square Coefficient of variation
- LOCALMAD local Median Absolute Deviation (MAD)
Platform: |
Size: 18432 |
Author: ANN |
Hits:
Description: 设计一个MATLAB界面，用于图像处理，实现了灰度拉伸，K邻近均值滤波，FFT变换，同态滤波，边缘检测，图像分割，均衡化等-A MATLAB interface design for image processing, gray stretch, K neighboring mean filtering, FFT transform, homomorphic filtering, edge detection, image segmentation, equalization
Platform: |
Size: 15360 |
Author: xiaoyu |
Hits:
Description: KMEAN C#
In data mining, k-means clustering is a method of cluster analysis which aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean. This results in a partitioning of the data space into Voronoi cells.
The problem is computationally difficult (NP-hard), however there are efficient heuristic algorithms that are commonly employed and converge fast to a local optimum. These are usually similar to the expectation-maximization algorithm for mixtures of Gaussian distributions via an iterative refinement approach employed by both algorithms. Additionally, they both use cluster centers to model the data, however k-means clustering tends to find clusters of comparable spatial extent, while the expectation-maximization mechanism allows clusters to have different shapes.
Platform: |
Size: 2048 |
Author: Truong |
Hits:
Description: 文件可以用来产生正态白噪声序列x(k) ，计算X(k)的均值，均方值，方差，相关函数-The file can be used to produce normal white noise sequence x (k), calculate the mean value of X (k), the mean square value, variance, correlation function
Platform: |
Size: 1024 |
Author: 林星辰 |
Hits:
Description: In data mining, k-means clustering is a method of cluster analysis which aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean. This results in a partitioning of the data space into Voronoi cells.
The problem is computationally difficult (NP-hard), however there are efficient heuristic algorithms that are commonly employed and converge quickly to a local optimum. These are usually similar to the expectation-maximization algorithm for mixtures of Gaussian distributions via an iterative refinement approach employed by both algorithms. Additionally, they both use cluster centers to model the data, however k-means clustering tends to find clusters of comparable spatial extent, while the expectation-maximization mechanism allows clusters to have different shapes.
Platform: |
Size: 1024 |
Author: pongpan |
Hits:
Description: 对于有缺测值的数据取平均
使用：
y=cmean(x,k,c)
x为包括有nan的数据，仪器观测中经常会出现
k为对x取第几维方向的平均
c的值可以控制缺测的条件，例如c取50，表示这一维方向上，如果有值的观测超过50个值才取平均，否则取nan；如果c取-5，表示缺测数在5以下时才做平均，否则返回nan-calculate the statistical mean of data but tick off nan
Usage
y=cmean(x,|k) or y=cmean(x,k,|c), k can be []
when [m,n,...]=size(x)
if k==1, [1,n,...]=size(y)
if not set k, then k=1, but if m==1, k=k+1,>>...
if k==2, [m,1,...]=size(y)
if k>2, [m,n,...,1(k demension),...]=size(y).
you can squeeze(y) by yourself to remove the dim(whose size(y,dim)==1)
c can control output bo be nan or mean, c =0 by default(namely, not work)
if c>0, if valuable numbers(notnan) < c, then output nan instead of mean
if c<0, if nan numbers >= c, then output nan instead of mean
Platform: |
Size: 1024 |
Author: zhoudg |
Hits:
Description: 利用matlab实现下面的问题：其中u(k)和z(k)分别为模型的输入和输出变量；v(k)为零均值、方差为1、服从正态分布的白噪声；为噪声的标准差；利用得到的u(k)和z(k)来辨识模型的阶次和参数。-Using matlab to achieve the following questions: where u (k) and z (k) respectively, the input and output variables of the model v (k) with zero mean and variance 1, follow a normal white noise noise standards Poor utilization u (k) and z (k) to identify the order and parameters of the model.
Platform: |
Size: 2048 |
Author: |
Hits:
Description: kmeans methode (k-means clustering is a method of cluster analysis which aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean)
Platform: |
Size: 2048 |
Author: |
Hits:
Description: K均值用于图像分割。
图像分割依据不同的颜色块进行聚类，对于相同颜色但深浅不同的目标可以进一步进行聚类然后分割。-The mean K used in image segmentation
Platform: |
Size: 133120 |
Author: 汪泉霖 |
Hits: