mse

几种分类器，包括adaboost和随机森林(Several classifiers including adaboost and random forest)

本例中通过使用MATLAB把声音信号进行分析，把有声音的部分进行过滤，分别过滤后 再把3个信号连在一起，使3个单独的声音信号转换为一个单独的连续声音信号。(In this case through the use of MATLAB to analyze acoustic signals, the voice has some filtering, respectively, after filtering the three signals together, so that three separate voice signal is converted to a single continuous sound signal.)

显示梯度图像，寻找图像中最大像素值并找出边界。(show the gradient of a picture,find the edge of big pixel in a picture)

将断层剖面划分为一组网格，并以固定的事件数或固定的扫描半径，计算每个网格点处的b值。然后用色彩表示断层面上b值的分布情况，从而反应断层的应力积累情况，便于地震学家分析该断层是否存在发震的可能(This subroutine creates a grid with spacing dx,dy (in degreees). The size will be selected interactively or grids the entire cross section. The b-value in each volume is computed around a grid point containing ni earthquakes or in between a certain radius. The standard deviation is calcualted either with the max. likelihood or by bootstrapping, when that box is checked. If not, both options can be assigned by the additional run assignment. Standard deviation of b-value in non-bootstrapping case is calculated from Aki-formula!)

Normalization from irregualr points

code to calculate Root Mean Square Error and Mean Square Skill Score

说明：  内含PDF文档有效的解释了三维重构的基本原理以及可以基于Matlab导成三维有限元模型(The PDF document effectively explains the basic principle of three-dimensional reconstruction and can be derived into three-dimensional finite element model based on matlab.)

说明：  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.)

Matlab source code to calculate temporal activity.

PRTool 5,matlab模式识别工具箱-PRTool 5 ,A Matlab Toolbox for Pattern Recognition(PRTool 5, matlab Pattern Recognition Toolbox-PRTool 5, A Matlab Toolbox for Pattern Recognition)