SOM_PREDICTION

《统计与自适应信号处理》中的程序，music源程序(err)

说明：  《MATLAB 7.0编程基础》第8章 （源码实例）主要讲解MATLAB图形用户界面设计("MATLAB 7.0 programming foundation" Chapter 8 (source code examples) mainly on MATLAB graphical user interface design)

GM(1,1)灰色预测函数，自动返回预测值，并包括残差，关联度，后验差检验。(GM (1,1) Grey forecasting function, automatic return predictive value, and includes the residuals, correlation, after test.)

文章的代码主要讲述椭圆型方程五点差分格式。用MATLAB软件来实现，大大提高了算法的精确度。(The code article focuses on five difference scheme for elliptic equations are considered. Use MATLAB software to realize, greatly improve the precision of the algorithm.)

matlab求解微分方程讲义，求解各类方程，介绍的很全面。(matlab solve differential equations lecture notes, solving all kinds of equations, introduced very comprehensive.)

用蒙特卡洛模拟的方法计算美式期权的价格以及基本描述,最小均方误差等算法的MSE的计算,采用波束成形技术的BER计算。( Monte Carlo simulation method of calculating the American option price and basic description, Minimum mean square error MSE calculation algorithm, By applying the beam forming technology of BE.)

一个基于阈值的粒子比较准则，用于处理多目标约束优化问题,该准则可以保留一部分序值较小且约束违反度在允许范围内的不可行解微粒,从而达到由不可行解向可行解进化的目的；一个新的拥挤度函数,使得位于稀疏区域和Pareto前沿边界附近的点有较大的拥挤度函数值,从而被选择上的概率也较大 从而构成解决多目标约束优化问题的混合粒子群算法。(A comparison based on the threshold criteria for the particle to handle multi-objective constrained optimization problem, part of the guidelines can keep order value is small and the constraint violation in allowing infeasible solutions within the particles, so as to achieve a feasible solution to the infeasible solution The purpose of evolution a new function of crowding, making the Pareto front in the sparse region and near the border points have a greater value of the congestion function, which is the larger the probability of selection so as to constitute solve multi-objective constrained optimization problems hybrid particle swarm algorithm.)

卡尔曼滤波的理解，利用一个简单的正弦信号来进行卡尔曼滤波，从而更好的理解卡尔曼滤波。(Understanding of the Kalman filter, with a simple sinusoidal signal to the Kalman filter so as to better understand the Kalman filter.)

ce programme permet de calculer la transformer en ondelette sous matlab

结合KNN(最近邻)算法和SVM(支持向量机)算法的分类器，，原创(KNN+SVM classification recognition algorithms, absolute originality)