nn_fuzzy

fastICA algorithm,blind signal processing helpful

This Code was developed from the code tracking red objects. it could be used for tracking faces if the color of the background is not the same color as human skin. This code runs fine even when you use the webcam on PC

压缩感知应用于载波频率偏移的创新作品 ，很有学习和研究的价值，给人耳目一新。(compressed sensing for estimation of CFO)

code producing figure 1.12 of the book "informatioin theory,inference,and learning algorithm"

盲源分离中最为经典的不动点算法——FastICA,该算法简单，对初学者有很大的帮助(Blind Source Separation in the most classical fixed-point algorithm- FastICA, the algorithm is simple, for beginners there is a great help)

复杂网络下各种模型的源程序实现matlab(The complex network model of the source matlab)

说明：  在matlab开发环境中，提取图像的像素点，并保存(Development environment in matlab to extract the image pixels, and save)

说明：  LMS/F组合算法由Shao-jen Lin和J.G.Jarris提出，它计算量增加不多却同时兼有LMS和LMF算法的优点。即收敛速度和稳态性能比基本LMS算法好，稳定性比LMF算法好。(LMS/F combination algorithm by the Shao-jen Lin, and JGJarris proposed, it is increasing much more calculations with both LMS and the LMF of the method. Ie, the convergence speed and steady-state performance better than the basic LMS algorithm, stability, better than the LMF algorithm.)

维纳滤波其设计，假设一个点目标在x，y平面上绕单位圆做圆周运动，由于外界干扰，其运动轨迹发生了偏移。其中，x方向的干扰为均值为0，方差为0.05的高斯噪声；y方向干扰为均值为0，方差为0.06的高斯噪声。(Wiener filter is designed, assuming a point target in the x, y plane around the unit circle in a circular motion due to external interference, its trajectory shift occurred. Which, x the direction of the interference of mean 0 and variance of 0.05 Gaussian noise y direction of the disturbance is zero mean Gaussian noise variance 0.06.)

有较好的参考价值，详细画出了时域和频域的相关图，使用大量的有限元法求解偏微分方程。( There are good reference value, Correlation diagram shown in detail the time domain and frequency domain, Using a large number of finite element method to solve partial differential equations.)