voicebox

用matlab实现的对sin函数类型的数据进行预测，这是现代数字信号处理课堂中的实验。(Using matlab implementation of the sin function of types of data to predict, which is a modern digital signal processing the classroom experiment.)

说明：  Matlab实现的图像配准程序，运用互信息，PSO等进行了优化。 (Matlab to achieve image registration procedures, the use of mutual information, PSO, etc. were optimized.)

对有类别监督的数据进行先行判别分析用于降维 数据为结构体类型，返回数据也是降维后的结构体类型。如(a.X a.y)a.X是带有类别的数据，a.y是类别信息(Synopsis: model = lda(data) model = lda(data,new_dim) Description: This function is implementation of Linear Discriminant Analysis. The goal is to train the linear transform which maximizes ratio between between-class and within-class scatter matrix of projected data.)

different types of convolution

可以对已有的信号求其循环谱，信号可以通用的。(The cyclic spectrum of the existing signal can be obtained, and the signal can be universal.)

产生宽带QPSK信号，并通过rrc滤波，用于检验发射通路的EVM等指标(Produce wideband QPSK signals, and by rrc filter, used to test the emission path of EVM and other indicators)

Dijkstra通用算法，最近一段时间一直在搞图论的相关算法，发现Dijkstra是一个很好的单元最短路径算法，此程序很短，非常适合学习(Generic Dijkstra algorithm, the most recent period has been engaged in graph theory algorithms, Dijkstra is a good unit the shortest path algorithm, this procedure is very short, very suitable for learning)

雷达信号处理 阵列波束形成 最小均方误差 采样协方差矩阵求逆 Capon波束形成等等(Radar Signal Processing)

单基雷达的空时导向矢量方向图的Capon谱示意图，基本空时自适应处理的程序(Oriented single radar vector direction of the space-time diagram of Capon spectrum diagram basic space-time adaptive processing procedures)

对花卉的数据集分别通过“五折法”、随机产生训练样本、欧式平方距离、绝对值距离、契比雪夫距离和马氏距离进行数据集的识别。(Data sets, respectively, for flowers through the " half of Law" , randomly generated training samples, European squared distance, absolute distance, Chebyshev distance and Mahalanobis distance for data set identification.)