NC_LMS

ga优化神经网权值&阈值程序。为了方便，我把几个matlab的程序都放在txt文件里，使用时，对照把txt内容复制进matlab即可(ga optimization neural network weights & threshold procedures. For convenience, I have several matlab programs on the txt file, use, control the content of the txt you copied into matlab)

说明：  用Matlab实现灰度图像SIFT（尺度不变特征变换）特征点的提取(Grayscale images with Matlab SIFT (Scale Invariant Feature Transform) feature point extraction)

Cellular Learning Automata Simulation

说明：  低通滤波器 基于MATLAB的低通滤波器，Z变换实现(low-pass filter based on MATLAB low-pass filter, Z transform)

改程序包可以进行字旋转链和量子寄存器建模等。(this program package helps modeling spin chains, qubit registers, etc. according to nonrelativistic quantum mechanics and some novel ideas from quantum information theory.)

说明：  matlab的简单练习，对图像进行基本操作。课堂练习而已(matlab simple exercises, perform basic operations on the image. Workshop only)

noise cancellation programs by lms and nlms

计算源信号和噪音之间的频谱差，用于噪音估计等。(Calculates the average log-spectral distance between CLEAN and NOISY signals)

1、系统仅处于训练状态。训练序列为2cos(2π×f1×n/fs)， 干扰为sin(2π×f2×n/fs)；序列长度为100个样点。 其中n=0…99，f1=1kHz，f2=2kHz，fs=8kHz。 2、采用LMS算法，均衡器输入为训练序列加上干扰，均衡器阶数为21阶，步长因子β=0.01。 3、在Matlab中编写浮点LMS算法，绘制均衡器的输出波形，并与训练序列进行对比。 (1, the system only in the training status. The training sequence is 2cos (2πf1n/fs), interference sin (2πf2n/fs) sequence length of 100 samples. Where n = 0 ... 99, f1 = 1kHz, f2 = 2kHz, fs = 8kHz. 2, using LMS algorithm, the input to the equalizer training sequence plus interference equalizer order is 21 bands, step factor β = 0.01. 3, the preparation of the floating-point LMS algorithm in Matlab, draw the output waveform equalizer, and with the training sequence comparison.)

说明：  这个文件中包含光谱数据筛选的程序以及应用例子，可以筛选光谱数据中的特征变量(This file contains the program and application examples of spectral data filtering, which can filter the characteristic variables in the spectral data)