t-copula

对图像进行模糊C均值算法聚类的一个源程序，可以学习一下算法。(The image fuzzy C-means clustering algorithm, a source, can learn about algorithms.)

说明：  Lstm进行时间序列预测，预测股票数据，按日的数据(Prediction of time series by LSTM)

MARKOv的应用 Step 1，求出各食品数据的增长率； Step 2，编程，使快速上升（>=0.03） ，在程序中用“2”表示；缓慢上升 (0,0.03)，在程序中用“1”表示；相对不变 0，在程序中用“0”表示；缓慢下降(-0.03,0) ，在程序中用“-1”表示；快速下降 (<=-0.03)，在程序中用“-2”表示；得到各类食品在不同时期的 ； Step 3，统计相邻增长率变化的一致的次数 ，就菜籽油而言，在2010-3-5到2011-3-15，其增长率为先快速增长后相对不变出现的次数为‘20’=1；具体42中食品的输出情况见附录； Step 4，对每一种食品，分五类进行叠加求和，分别为“先快速增长”、“先缓慢增长”、“先相对不变”、“先缓慢下降”、“先快速下降”；得到 ； Step 5，求出一步转移概率矩阵E ；即 每行的每五个数据依次与对应的 相除，得到概率矩阵E 的一行； Step 6，使用求出的一步转移概率矩阵E ，并求出不同时期的状态概率(MARKOv application Step 1, to obtain the growth rate of each food data Step 2, the programming, so fast rise (> = 0.03), in the program with " 2" slowly rising (0,0.03), used in the program " 1" indicates relatively constant 0, by the program " 0" indicates slow decline (-0.03,0), in the program by " -1" rapid decline (< =- 0.03), in the program with " -2" means get all kinds of food at different times Step 3, the same number of adjacent growth statistics change on rapeseed oil, in 2010-3-5 to 2011-3-15, Its growth rate is relatively rapid growth of the number of times after the first change occurs is ' 20' = 1 42 Output of specific food in Appendix Step 4, for each kind of food, five points are superimposed sum, respectively, " First of rapid growth," " first slow growth" , " first relatively unchanged" , " first slow down" , " first rapid decline" get Step 5, to obtain a step)

metodo falsa posicion ecuaciones soluciones aproximaciones codigo

数值计算，微分和积分的几个例子，针对于MATLAB2011a(numerical calculation)

系统深入得多，一些在第一本中只略微论及的高级内容，这里讲得完整透彻。理论与实践并重，可操作性强。例题都有可直接运行的matlab文件，正文中也给出了关键部份方便阅读。读完这本书，住真水平可以上一个台阶，基本上可以达到系统级仿真的能力。(System is much more depth, some in the first book deals with only a slight premium content here put a complete and thorough. Theoretical and a practical, easy to operate. Example are directly run matlab file, the text also gives a key part of user-friendly. After reading this book, the level of living can be really a new level, but basically can achieve system-level simulation capabilities.)

特征选择。。。。。。。。。。。。。。。。。。(feature selection。。。。。。。。)

用Baum-Welch算法来迭代估计一个隐马尔科夫模型（HMM）的初始状态概率分布以及其状态转移概率矩阵。其中文件有mainfile_B_W.m为主函数，Baum_Welch.m为Baum-Welch算法迭代函数，Forward_variable.m与Backward_variable.m与Gamma_variable.m与Ksi_variable.m是需要计算的四种因子，B_pdf.m为混淆散射概率密度函数。(It s Baum-Welch algorithm for iteratively estimating the initial states probability distribution and state transition probability matrix. In the folder, "mainfile_B_W.m" is the main function, and "Baum_Welch.m" is the iteration Baum-Welch function. Also these four file "Forward_variable.m " "Backward_variable.m" "Gamma_variable.m" "Ksi_variable.m" are used for calculating crucial factors. "B_pdf.m" is the emission probability distrubution function.)

The Rice Wavelet Toolbox (RWT) is a collection of Matlab M-files and C MEX-files for 1D and 2D wavelet and filter bank design, analysis, and processing. The toolbox provides tools for denoising and interfaces directly with our Matlab code for wavelet domain hidden Markov models and wavelet regularized deconvolution.

past-music算法，用来进行谐波分量的求解。速度快，效果佳(past-music algorithm used for solving the harmonic components. Fast, good effect)