A-variable-object-least-squares

一种无监督的数据离散化方法，程序简单，运行时间短，效果比较显著(An unsupervised data discretization methods, procedures easy, run a short time, compared the effect of significantly)

说明：  给定栅格图象及其变形后的栅格图象，请你首先确定各栅格交叉点的位移，然后采用控制栅格的方法编写实现图象几何校正的程序(A given image and deformed grid raster images, you first determine the displacement of each grid intersection, and then using the control method of preparation of raster image geometric correction procedures to achieve)

说明：  基于压缩感知的的图像去噪，用于不同维数，(image denoising based on cs)

Image edge detection ACO

DahlUrsin, modeling amplitude versus offset

说明：  MATLAB编程教程，对初学者非常有用。 请大家多用用MATLAB编程(MATLAB programming tutorial, very useful for beginners. Please keep on with the use of MATLAB programming)

回归分析的好东西spss09回归分析 (Regression analysis regression analysis of the good things spss09 )

卡尔曼滤波算法 包括数据、程序 已经打包(Kakman filter)

solutions to linear systems numerical analysis

EM 算法是Dempster，Laind，Rubin于1977年提出的求参数极大似然估计的一种方法，它可以从非完整数据集中对参数进行 MLE估计，是一种非常简单实用的学习算法。这种方法可以广泛地应用于处理缺损数据、截尾数据以及带有噪声等所谓的不完全数据，可以具体来说，我们可以利用EM算法来填充样本中的缺失数据、发现隐藏变量的值、估计HMM中的参数、估计有限混合分布中的参数以及可以进行无监督聚类等等。(EM algorithm is a method for maximum likelihood estimation of parameters proposed by Dempster, Laind and Rubin in 1977. It can estimate parameters from incomplete data sets by MLE. It is a very simple and practical learning algorithm. This method can be widely used in dealing with the so-called incomplete data such as defective data, censored data and noisy data. Specifically, we can use EM algorithm to fill missing data in samples, find hidden variable values, estimate parameters in HMM, estimate parameters in finite mixed distribution, and so on. Unsupervised clustering and so on.)