EM-GMM

灰色预测模型称为CM模型，G为grey的第一个字母，M为model的第一个字母。GM（1，1）表示一阶的，一个变量的微分方程型预测模型。GM（1，1）是一阶单序列的线性动态模型，主要用于时间序列预测。 一、GM（1，1）建模 设有数列 共有 个观察值 对 作累加生成，得到新的数列 ，其元素 （5-1） 有： 对数列 ，可建立预测模型的白化形式方程， （5-2） 式中： ——为待估计参数。分别称为发展灰数和内生控制灰数。设 为待估计参数向量 则 按最小二乘法求解， 有： （5-3） 式中： （5-4） （5-5） 将（5-3）式求得的 代入（5-2）式，并解微分方程，有 （1，1）预测模型为： （5-6）(gray forecasting model called CM model, the G-gray of a letter, the M model for the first letter. GM (1,1), a band of a variable type of differential equation models. GM (1,1) is a sequence of single-band linear dynamic model, mainly for time series prediction. A GM (1,1) model series with a total value of observation for the cumulative production, to a new series of its elements (5-1) : The series, we can establish the prediction model albino form of equation (5-2) where :-- to question the estimated parameters. The development will be known as the gray and hygiene control within a few gray. Set parameters to be estimated according to Vector least squares method, are : (5-3) where : (5-4) (5-5) (5-3) that obtained in lieu of income (5-2)- and solutions differential equations, (1 1) Fore)

name : Simulation Tools for Wireless Sensor Networks E. Egea-López, J. Vales-Alonso, A. S. Martínez-Sala, P. Pavón-Mariñ o, J. García-Haro Department of Information Technologies and Communications Polytechnic University of Cartagena, Spain. Email: {esteban.egea, javier.vales, alejandros.martinez, pablo.pavon, joang.haro}@upct.es  Corresponding author. Address: Campus Muralla del Mar, 30202, Cartagena, Spain. phone: +34 968 325314, fax: +34 968 325973, e-mail: joang.haro@upct.es

说明：  带约束算子的粒子群算法，用于matlab环境，对于搞研究的教师有很大帮助(Operator Constrained Particle Swarm Optimization for the matlab environment, engage in research for teachers are very helpful.)

说明：  在matlab中运行的用于对数据进行正交信号校正的程序(Matlab running in the data for orthogonal signal correction procedure)

将年月日时分秒转换为某一基准年开始的总秒数(The date(YYYY，MM，DD，hh，mm，ss) is converted to a base year of the total number of seconds.)

Four crucial issues are considered by the proposed HoAL: 1) unlike binary cases, the selection granularity for multilabel active learning need to be fined from example to examplelabel pair 2) different labels are seldom independent, and label correlations provide critical information for efficient learning 3) in addition to pair-wise label correlations, high-order label correlations are also informative for multilabel active learning and 4) since the number of label combinations increases exponentially with respect to the number of labels, an efficient mining method is required to discover informative label correlations

利用最小二乘法进行拟合多元非线性方程，用于特征降维，特征融合，相关分析等,基于chebyshev的水声信号分析。( Multivariate least squares fitting method of nonlinear equations, For feature reduction, feature fusion, correlation analysis, Based chebyshev underwater acoustic signal analysis.)

说明：  本程序是用来预测简单的时间序列的数据，欢迎大家一起来探讨学习(time series data of the prediction)

matlab receiver for image transmission using usrp

用于盲信号分离的原代码，可以根据输入源信号的基本统计特征,由观测数据进行信号分离,最终恢复出源信号。 (Blind signal processing has become an emerging subject in signal processing in these years, which depends on the source signal statistical characteristics to separate the signals from the observation data and recovers the source signals in the end. )