搜索资源列表
javasom_1[1].0.0
- Kohonen网络的学习过程可描述为：对于每一个网络的输入，只调整一部分权值，使权向量更接近或更偏离输入矢量，这一调整过程就是竞争学习。随着不断的学习过程，所有输入矢量都在输入矢量空间相互分离，形成了各自代表输入空间的一类模式，这就是Kohonen网络的特征自动识别的聚类功能。请解压缩后按照readme提示进行操作。-Kohonen network learning process can be described as follows : for each one network input,
mmse_diedai08
- 最佳权向量自适应迭代更新MSE对比波形，内有详细说明，供大家参考。-optimal weight vector adaptive waveform update MSE contrast, with detailed notes for reference.
run_qrd_lsl
- ％ because we do not truncate and shift the convolved input ％ sequence, the delay of the desired output sequence wrt ％ the convolved input sequence need only be the delay ％ introduced by the ideal weight vector centred at n=5
weightpercent
- Returns weighted percentiles of a sample given the weight vector w ％ The idea is to give more emphasis in some examples of data as compared to ％ others by giving more weight. For example, we could give lower weights to ％ the outliers. ％ The
AHP.rar
- 层次分析法，用matlab编制，输入任何矩阵后，可进行一致性检验，标准化权重向量，最大特征值计算。运行在通过matlab 2008a通过。,AHP using matlab prepared to enter any matrix may conformance testing, standardization of weight vector, the largest eigenvalue calculation. Running through the adoption of matlab
Perceptron-and-ADALINE-network
- 这些程序包括以下方面1.使用感知器和ADALINE网络对字母进行识别。2.随机选取初始权向量，选取适当的迭代步长(对ADALINE网络)，用给出的四个输入训练样本，对上述两个网络分别进行训练，直到网络收敛；3.对Adaline网络选取不同的值，分别画出误差曲线，观察它们的变化规律；4.对感知器选取不同的初始权向量，分别计算每类训练样本到超平面的距离，观察它们的异同；5.训练结束后，检验网络的识别能力(使用100个检测样本，对应于每个取25个含噪的变形)：6.比较Adaline和单神经元感知器的分
fixed_increment_method
- 该程序可实现将样本集转化为增广模式向量，固定增量法求解权向量。-The program can be realized sample set will be translated into augmented model vector, fixed incremental weight vector method.
sample9
- 绘制互相关信号,采用权向量和迭代次数求滤波器系数.-Drawing cross-correlation signal, weight vector and the iteration number for filter coefficients.
wirelesscomm
- In this project we analyze and design the minimum mean-square error (MMSE) multiuser receiver for uniformly quantized synchronous code division multiple access (CDMA) signals in additive white Gaussian noise (AWGN) channels.This project is mainly bas
UCA_SMI
- QR分解SMI算法的目的正是要避免直接来解线性方程，而是将自相关矩阵分解，并利用Givens旋转实现数据矩阵的QR分解，最终将权矢量的求解问题转化为三角线性方程组的求解问题。-QR decomposition SMI algorithm, whose objective it is to avoid directly to solution of linear equations, but will auto-correlation matrix decomposition, and the u
beampattern
- 由切比雪夫权值矢量及阵元间距画波束方向图：有x输入，计算该方向的值,否则计算所有方向的值，画出方向图-By the weight vector and the array element spacing beam pattern draw: there is x input, calculate the value of the direction, otherwise the value calculated in all directions, draw the direction of g
NNFunctions
- NN Functions a program in Lisp to demonstrate working of an artificial neuron. (Enter an input vector X and weight vector W. Calculate weighted sum XW. Transform this using signal or activation functions like logistic, threshold, hyperbolic-tangent
rls
- 自适应rls的算法仿真，最后图包括学习曲线，误差权矢量-Rls adaptive algorithm simulation, the final plan, including the learning curve, the error weight vector
1
- LMS在加性高斯白噪声信道下迭代2000步的输出能量和误差权矢量-LMS in the additive white Gaussian noise channel output of the next iteration step 2000, the energy and the error weight vector
1
- LMS算法误差权矢量曲线盲多户检测中的应用-LMS algorithm weight vector error curve Blind detection of multi-family
BP
- 这是一个基于matlab的三层神经网络算法，赋予不同的向量可对权值进行训练-This is a three-layer neural network matlab-based algorithm can give different weight vector for training
som(Jal.You)
- SOM神经网络（自组织特征映射神经网络）是一种无导师神经网路。网络的拓扑结构是由一个输入层与一个输出层构成。输入层的节点数即为输入样本的维数，其中每一节点代表输入样本中的一个分量。输出层节点排列结构是二维阵列。输入层X中的每个节点均与输出层Y每个神经元节点通过一权值(权矢量为W)相连接，这样每个输出层节点均对应于一个连接权矢量。 自组织特征映射的基本原理是，当某类模式输入时，其输出层某一节点得到最大刺激而获胜，获胜节点周围的一些节点因侧向作用也受到较大刺激。这时网络进行一次学习操作，获胜节点
SMI
- 块自适应处理算法先有采样快拍数据计算采样协方差矩阵，再来计算自适应权矢量。典型的块自适应处理算法为采样矩阵求逆(SMI)算法。仿真了采样矩阵求逆法(SMI)的波束形成方向图旁瓣的高低受信号快拍数的影响。-Block adaptive processing algorithms prior sample snapshot data of the sampling covariance matrix, again computing adaptive weight vector. A typical
SMI
- SMI方向图和性能(波束形成算法，自己修改） 优点： 收敛速度快 缺点：1 当阵元输出含有较强的期望信时，或者期望信号与干扰信号相关时，性能急剧下降.2 由于权向量含有方向矢量，因此对阵列的幅相差非常敏感；3 ,期望信号的功率不能过大，比干扰信号相差几十dB，也就是说，在小期望信号、大干扰信号情况下，也可进行SMI处理。 -SMI pattern and performance (beamforming algorithm, make changes to) Advan
vector-optimization
- 针对标准的Capon 波束形成在阵列导向矢量存在误差时出现严重性能退化, 本文从权矢量模约束和不确定集两个角度出发, 提出矢量最优化稳健波束形成方法, 并通过二阶锥规划进行有效求解.-When there exists the array steering vector error , the performance of standard Capon beamforming will be severely degraded .In this paper , in the view of