BP_recog_hao

顺序聚类算法的改进-MBSAS，并在MATLAB上实现该算法。(Clustering algorithm in order to improve-MBSAS, and the algorithm in MATLAB to achieve.)

LMNN,监督距离度量学习目前效果最好的(lmnn)

采用非线性动力学分析方法来研究互联网拥塞控制系统中存在 的复杂非线性动力学现象，如分岔(bifurcation)、混沌(chaos)等，针对不同参数该组源程序仿真互联网拥塞控制系统频闪模型的非线性动力学现象。(The programs study the dynamic behavior in stroboscopic model of Internet congestion control system is studied. )

edge detection sobel

说明：  利用matlab的gui设计的带有输入和输出功能的界面图形，实现等距线可视性很强，自己独立编写的数据和图形，学习非常值得,实现pca压缩图像，matlab语言(The interface graphics with input and output function designed by GUI of Matlab can realize the visibility of isometric line. The data and graphics compiled independently by oneself are worth learning. The realization of PCA compressed image and MATLAB language is very worthwhile.)

Contain a function to delete outliers from a serie. Only need a column vector to start. Can add a alpha parameter to configure the way to delete outliers

设计一个IIR数字低通滤波器， 逼近一组模拟滤波器的指标参数（通带截止频率Wp=2*pi*2000rad/s,阻带边界频率Ws=2*pi*3000rad/s,通带波纹 Rp=3db, 阻带衰减Rs=15db, 采样频率f=10000Hz）； 分别用脉冲响应不变法和双线性变换法实现设计，列出传递函数并描绘模拟和数字滤波器的幅频和相频响应曲线。用上述设计滤波器完成几组给定信号的滤波，证明滤波器的有效性和滤波范围限制.(Design a IIR digital low-pass filter, approaching a group of analog filters indicator parameters (pass-band cut-off frequency Wp = 2* pi* 2000rad/s, stop-band edge frequency Ws = 2* pi* 3000rad/s, pass-band ripple Rp = 3db, stop-band attenuation Rs = 15db, the sampling frequency f = 10000Hz) were used to change the impulse response method and bilinear transformation method for design, list and describe the transfer function of analog and digital filter amplitude-frequency and phase-frequency response curve. A filter was designed with the completion of the above-mentioned groups of a given signal filtering, to prove the effectiveness of filters and filter limit)

说明：  caiyang 演示采样定理，时域采样，频谱周期延拓(caiyang demonstration sampling theorem, time domain, frequency cycle Extension)

MATLAB实用程序(MATLAB utility SSSSSSSSSSSSSSSSSSSSSSSSSS )

Function H for Method upwind, Lax_Friedrichs and Lax_Wendroff for equation of advection 1d in Matlab