BM3D_v1.7.6_x64

k means techniques de classification non supervisée (clustering) les plus utilisées. Etant donné un entier K, K-means partitionne les données en K groupes, ou "clusters", ou "classes" ne se chevauchant pas. Ce résultat est obtenu en positionnant K "prototypes", ou "centroï des" dans les régions de l espace les plus peuplées. Chaque observation est alors affectée au prototype le plus proche (règle dite "de la Distance Minimale"). Chaque classe contient donc les observations qui sont plus proches d un certain prototype que de tout autre prototype

A small Blind Source Separation example

说明：  切普信号生成，可以改变信号的起始频率和最终频率。(Be able to generate a linear-FM “chirp” signal that sweeps in frequency fromω1 = 2pi*f1 to ω1 = 2pi*f2 as time goes from t = 0 to T2. You can change any of these values.)

— In ultra-wideband (UWB) indoor dense multi-path transmission environments, traditional coherent Rake receiver requires estimation of channel parameters. Moreover, in order to collect enough signal energy from dense multi-path components, the number of Rake fingers must be large enough. Consequently, the receiver structure is too complex to be implemented in practice.

LS AND RLS implementation

说明：  采用传统的遗传算法，结合帕累托策略，同时考虑环境和经济性。(The traditional genetic algorithm, combined with Pareto strategy, is used to consider both environment and economy.)

SIMULASI3.ZIP SIMULASI3 ALAGORITHM.............. 55

一维球坐标系下辐射传递方程切比雪夫配置点谱方法求解(One-dimensional spherical coordinates radiative transfer equation Chebyshev collocation spectral method for solving)

使用Matlab实现的关于主分量分析PCA的方法的案例，可先实现仿真(Matlab achieved using principal component analysis PCA on the case method, the simulation can be achieved)

此程序基于Matlab提升法实现第二代提升小波变换(This program is based on the second generation of lifting wavelet transform Matlab enhance law)