FastICA_25

说明：  某厂生产甲乙两种口味的饮料,每百箱甲饮料需用原料6千克,工人10名,可获利10万元 每百箱乙饮料需用原料5千克,工人20名,可获利9万元.今工厂共有原料60千克,工人150名,又由于其他条件所限甲饮料产量不超过8百箱.问如何安排生产计划,即两种饮料各生产多少使获利最大.进一步讨论: 1)若投资0.8万元可增加原料1千克,问应否作这项投资. 2)若每百箱甲饮料获利可增加1万元,问应否改变生产计划. (Factory production of beverages and B in two flavors, each box 100 need a drink six kilograms of raw materials, workers 10 to 100,000 yuan profit B beverage cartons per cent five kilograms of raw materials needed, workers 20, the profitability of 90,000 yuan. this plant a total of 60 kilograms of raw materials, 150 workers, and because other conditions do not yield a beverage cartons more than 8 per cent. ask what arrangements the production of projects, namely the production of two the number of drinks to make the biggest profit. further discussion : 1) if the investment 08,000 yuan to increase one kilograms of raw materials and asked whether we should make this investment. 2) If a beverage cartons per cent profit increase 10,000 yuan and asked whether or not to change the production plan.)

matlab，最小拍控制系统，输入为速度信号，输出控制器输出变量有波纹。有波纹最小拍控制系统(matlab, the smallest shot control system, the input speed signal, the output controller output variable corrugated. Corrugated minimum Beat Control System)

多输入多输出系统的状态空间simulink模型，可以直接在里面修改运行获得阶跃响应(MIMO viarieble system with state space model)

用于计算两个一维向量的互信息,information.m用于计算互信息值，需要调用histogram2.m(Use for calculating the mutual information with two one-dimensional vector.)

用于基于图像纹理特征提取的图像检索，其中包括matlab中自带的graycomatrix和graycoprops两个m文件，和四个用于提取图像纹理特征及图像检索的m文件(For image retrieval, including matlab comes graycomatrix, and graycoprops two m files, and four m-files used to extract image texture features and image retrieval based on image texture features extracted)

改进FAST-ICA采用3阶牛顿进行迭代，收敛速度更快(Improvements of the FAST-ICA order 3 Newton iterations, convergence is faster)

说明：  对结构损伤加速度值进行小波分析，再对其加载噪声，最后进行去噪处理(Acceleration values for structural damage analysis, and then the noise of its load, and finally de-noising)

《滑模变结构控制MATLAB仿真》里的源代码，对学习滑模变技术相当有用！(" MATLAB Simulation of sliding mode control" in the source code, the sliding mode technique is useful to learn!)

This is a reference implementation for the synthetic experiments on lower linear envelope inference and learning described in "Max-margin Learning for Lower Linear Envelope Potentials in Binary Markov Random Fields", Stephen Gould, ICML 2011.

是我见过讲的最详细的关于稀疏信号的书，讲解了稀疏分解，以及书后的程序都是很好用的(Ive seen talking about the most detailed book about the sparse signal, explained the sparse decomposition, and program books are well used after)