aeroblk_HL20

Text to Speech in MATLAB

matlab实现的meanshift算法，本人使用效果不错(meanshift on matlab )

说明：  自适应滤波器。自适应滤波器为11个权系数的FIR结构。（1）不同的方差σ2（2）LMS算法画出一次实验的误差平方收敛曲线，训练长度为500，给出滤波器系数；进行20次独立实验，给出平均收敛曲线。不同步长值的比较。（3）RLS算法，LMS和RLS算法的比较 (Adaptive filter. Adaptive filter weights for 11 of the FIR structure. (1) different variance σ2 (2) LMS algorithm for an experiment to draw square error convergence curve, the training length is 500, given filter coefficient conduct 20 independent experiments, given the average convergence curve. Are not synchronized long value comparison. (3) RLS algorithm, LMS and RLS Algorithms)

支持向量机是用于分类与回归的技术。由于其自身的诸多缺点，如无法获得概率输出，需要估计一个误 差参数C，以及必须使用Mercer 核函数等。相关向量机算法，克服了SVM 上述缺点，RVM 能获得与SVM 相比拟的推 广性能，并且更为稀疏。在此基础上，文中介绍了一种RVM 回归用于分类的新分类方法，用RVRC 来表示。并通过 实验证明了它的可行性。(Support vector machines for classification and regression techniques. Due to its many shortcomings, such as the probability that the output can not be obtained, need to estimate an error parameter C, and the need to use Mercer kernel function and so on. Relevance vector machine algorithm to overcome these shortcomings of SVM, RVM and SVM can be obtained comparable to promote performance and more sparse. On this basis, the paper introduces a RVM regression was used to classify new classification method, RVRC to represent. And through experiments proved its feasibility.)

求一些信号的均方根时延扩展、信道衰弱速率(Seeking root mean square delay spread of some signals)

un article de J.Betz contient la modulation BOC

这是一个滤波器，用来将时域信号转换位频域进行分析滤波后，再转为时域信号的小程序(This is a filter for converting the digital time domain signal analysis of the frequency domain filtering, a time domain signal of the small sub-program)

Oustaloup approximation for FOC

贝叶斯程序。。。。。。。。。。。。。贝叶斯程序贝叶斯程序(NaiveBayes1.m)

前面搞错了,把五个文件一起打包上传了,现在一个一个传 预加重算法,与直接传输 算法比较 模拟程序 程序头上有详细的注释和使用说明(Made a mistake in front of the pack with five file upload, and now a mass of a pre-emphasis algorithm, compared with the direct transmission algorithm simulation program procedures detailed head notes and use)