qlearning_example

基于仿射变换的数字图象置乱技术, MATLAB源程序代码.(The digital image scrambling technology based on affine transformation MATLAB source code )

利用自然梯度算法，采用的是通用的平面波展开法,可以得到很精确的幅值、频率、相位估计。( Use of natural gradient algorithm, Using common plane wave expansion method, You can get a very accurate amplitude, frequency, phase estimation.)

music高阶谱分析算法，基于kaiser窗的双谱线插值FFT谐波分析，FMCW调频连续波雷达的测距测角。( music higher order spectral analysis algorithm, Dual-line interpolation FFT harmonic analysis kaiser windows, FMCW frequency modulated continuous wave radar range and angular measurements.)

说明：  这次上传的程序源码是关于EM算法的编程实例。(The upload process is on the EM algorithm source code programming examples.)

Self devrloped Demos for beginners of Matlab generate HTML report and enjoy

基于内容的图像识别特征提取部分——k-均值聚类分割获取形状等信息(Content-based image recognition, feature extraction part of- k-means clustering segmentation information for shape)

在运动的位置叠加噪声。进行JPDA概率数据关联及kalman滤波。 两运动目标在x-y平面做匀速直线运动。初始位置是（4000,1200）（300,1500）速度分别是（200,200）（400,200）传感器对量目标进行位置状态量测。 采样间隔T=1,点数n=80.检测概率为1，正确量测落入跟踪内的概率为0.99，杂波均匀分布的密度为2个/km2由RAND函数产生在[0，1]上均匀分布的随机变量，跟踪门限为9.21。 (Superimposed noise in the position of the movement. JPDA probabilistic data association and kalman filtering. Two moving targets uniform linear motion in the xy plane. The initial position (4000,1200) (300,1500) speed (200,200) (400,200) position sensor on the amount of target state measurements. Sampling interval T = 1, points n = 80. Detection probability of correctly measured fall into the tracking probability 0.99, 2/km2 clutter uniform distribution of density generated by the RAND function [0,1] uniformly distributed random variables tracking threshold of 9.21.)

说明：  actif power filter with inductance and capasitor

MACD量化趋势交易的MATLAB编程，回测各指标的代码以及Excel输出(MACD quantify trends transaction MATLAB programming, backtesting of each index code, and Excel output)

基于强跟踪滤波算法的惯性加GPS组合导航matlab仿真程序(Strong tracking filter algorithm based IMU+ GPS integrated navigation matlab simulation program)