Program_exercise3

利用双缓冲技术实现使得能用鼠标在桌面上轻松画取圆形，(The use of double-buffering technology makes the mouse can be used to easily draw on the desktop to take round,)

memcached的C,c++实现，使用它可以方便的实现跨进程数据共享(memcached s C, c++ realize, it can be easily achieved using inter-process data sharing)

自编奇异值MATLAB程序，利用奇异值去除信号噪声。以供参考。(Editing the singular value MATLAB, using the singular value to remove noise signal. For your reference.)

<最优化及matlab的应用>课后题的编程(Most optimized operating)

代码检查的PC-lint工具源码,实现在Windows上图形化检测代码是否符合规范。(the source code of PC-lint on windows)

在EFI架構下寫一隻PPI sample(PEIM to PEIM Interface),使兩隻PEIM可透過此PPI互相傳遞資料. Write a PPI sample to make two PEIMs able to communicate each other.(Write a PPI sample to make two PEIMs able to communicate each other.)

说明：  EM 算法是Dempster，Laind，Rubin于1977年提出的求参数极大似然估计的一种方法，它可以从非完整数据集中对参数进行 MLE估计，是一种非常简单实用的学习算法。这种方法可以广泛地应用于处理缺损数据、截尾数据以及带有噪声等所谓的不完全数据，可以具体来说，我们可以利用EM算法来填充样本中的缺失数据、发现隐藏变量的值、估计HMM中的参数、估计有限混合分布中的参数以及可以进行无监督聚类等等。(EM algorithm is a method for maximum likelihood estimation of parameters proposed by Dempster, Laind and Rubin in 1977. It can estimate parameters from incomplete data sets by MLE. It is a very simple and practical learning algorithm. This method can be widely used in dealing with the so-called incomplete data such as defective data, censored data and noisy data. Specifically, we can use EM algorithm to fill missing data in samples, find hidden variable values, estimate parameters in HMM, estimate parameters in finite mixed distribution, and so on. Unsupervised clustering and so on.)

All graph algorithms thats helps to understand the concepts

该程序是灰色系统预测，可以对宏观经济进行分析，功能很强大，适应性良好，在matlab环境下运行。G(1,1)预测控制，最小二乘法(G（1,1）prediction control)

精通matlab综合辅导与指南。。用c++编程实现的matlab计算。。(Proficient in matlab General counseling and guidance. . Programming using c++ implementation of matlab calculation. .)