LMS

Matlab编写的粒子滤波算法源程序，内含中文程序注释(Matlab prepared on the particle filter algorithm simulation source)

很不错的图像分割matlab源代码，代码相对比较简单，易于理解(Very good image segmentation matlab source code, the code is relatively simple, easy to understand)

压缩感知方法演示，1-D信号压缩传感的实现(正交匹配追踪法Orthogonal Matching Pursuit) 测量数M>=K*log(N/K),K是稀疏度,N信号长度,可以近乎完全重构(1-D signal is compressed sensing to achieve (orthogonal matching pursuit method Orthogonal Matching Pursuit) number of measurements M> = K* log (N/K), K is the sparsity, N signal length, can be almost completely reconstructed)

说明：  一些matlab学习用资料和数据分析的相关资料，供学习python和数据分析的使用。(Some matlab learning materials and data analysis related materials are used for learning python and data analysis.)

Spectrum Sensing Technique

MUSIC算法是一种基于矩阵特征空间分解的方法。从几何角度讲，信号处理的观测空间可以分解为信号子空间和噪声子空间，显然这两个空间是正交的。信号子空间由阵列接收到的数据协方差矩阵中与信号对应的特征向量组成，噪声子空间则由协方差矩阵中所有最小特征值（噪声方差）对应的特征向量组成 (MUSIC algorithm is a feature space based on matrix decomposition method. From the geometric point of view, the observed signal space can be decomposed into signal subspace and noise subspace, it is clear that two spaces are orthogonal. Signal subspace to the data received by the array covariance matrix and the eigenvector corresponding to the signal component, the noise subspace from the covariance matrix of all the smallest eigenvalue (noise variance) eigenvector corresponding to the composition )

基于pid模糊控制的船舶转向控制 simulink(Ship steering control based on PID fuzzy control Simulink)

用costas环实现载波同步。假设发送信号载波频率为3.563Mhz，本地初始频率为3.5628Mhz。程序中直接给出了环路滤波器的参数C1,C2。(this is synchoronization on time recovery basing on matlab)

最优估计最有用的 程序的新方法 ，对于新人学习有很大的帮助，请大家多多指教 共同交流(Matable In order to improve the adaptation property of the system by making the best of the direction of distribution of radar measurement errors, a optimum estimation algorithm for Kepler orbit is proposed, which yields a maximum likelihood estimation of the Kepler orbit with multi-spatio-temporal state based on both the preprocessing of the small manoeuvring model and the determining of initial value parameter with two spatio-temporal state. )

MMT is a Matlab toolbox implementing the multi-task Lasso models, including: (i) the Lasso (ii) the standard multi-task Lasso, i.e. the group Lasso (iii) the structured input-output multi-task Lasso, a.k.a. the two-graph guided multi-task Lasso proposed in [1]. The last case (iii) subsumes the special cases: tree-guided and the feature-graph guided multi-task Lasso. The core optimization algorithm for solving this model is developed in C to enhance greater computational efficiency. In particular, current scalability of the coefficient matrix that has been tested for MMT is 104*104! The structured input-output multi-task Lasso model is well-suited for addressing the expression quantitative trait loci (eQTL) mapping problems which are of the intrinsic high-dimensional nature. Details can be found in [1].