MATLAB-

自己编的绘制koch曲线和serpinki垫的matlab小程序(Own drawing pad koch curve and the serpinki applet matlab)

说明：  M-file to calculate and plot the velocity and

semi conductor text book

针对灰度图像，提出一种基于知识的人脸检测方法。 提出了一种给予支持向量机的人脸检测方法。 提出了一种基于小波分解的LDA人脸识别方法。 提出了一种基于小波和DCT的人脸识别方法。 提出了一种机遇CEDT和支持向量机的人脸分类和识别方法。 (For gray-scale images, a knowledge-based face detection methods. A support vector machine method of face detection. A wavelet decomposition of the LDA-based face recognition methods. A wavelet and DCT-based face recognition methods. A CEDT opportunities and support vector machine classification and face recognition.)

利用matlab做一些基礎的矩陣運算指令，包含建立矩陣、計算及求特徵值等等.....(Using matlab to do some basic matrix operations command, including the establishment of matrix eigenvalue calculation and demand and so on .....)

利用matlab在图像中加入各种随机噪声，包括调用函数，以及randn使用、以及介绍send和state的区别。(Using matlab to join all kinds of random noise in the image, including the calling function, and randn use, as well as the introduction of the difference between send and state.)

ofdm信号通过sui信道，希望对大家学习有所借鉴学习(ofdm signal through sui channel, and they hope to learn from everyone has been learning to learn)

设计一个简单的GUI界面，对导入的音频文件进行时域特性分析。计算其时域波形、平均幅度、短时能量、短时过零率。对该音频文件进行加噪处理，分析其时域波形变化。设定一个固定时长，对该音频添加回声并保存为“*.wav”格式文件。 对音频文件的时域分析是对音频文件进行传输、降噪及其他再编辑的基础。其中短时能量可用作清浊音区分的标准和语音起始点的判断。短时过零率可用作判断有声段与无声段和从背景噪声中提取语音信号。 (Design a simple GUI interface that time domain characteristics analysis on import audio files.Calculate the average time domain waveform, amplitude, short-time energy and short-time zero crossing ratio.Add to deal with the noise of the audio files, analyze the changes in the time domain waveform.To set a fixed time, to add the audio echo and saved as a *. Wav format file. The time domain analysis of the audio file was transmitted to audio files, noise reduction and other editor again.The short-time energy voiced distinguish between standard and can be used as a clear voice starting point judgement.Short-time zero crossing ratio can be used as a judgment audio segment and silent segment and extract speech signal background noise.)

说明：  三次贝塞尔曲线拟合，拟合效果可控,matlab解释详细明了(Cubic Bezier curve fitting, the fitting effect is controllable, matlab explanation is clear)

VC维理论和结构风险最小化准则是统计学习理论中的重要内容,基于这一理论的支持向量机算法由于具有好的泛化性能受到重视,并被研究用于文本分类问题.基于多项式核的研究工作认为SVM的泛化能力不受多项式阶数的影响,并且能够处理很高维的分类问题,用于文本分类无需进行特征选择.研究发现,随着多项式核阶数的升高,SVM文本分类器会出现过学习现象,并且特征数越多越明显,特征选择是必需的.通过估计函数集的VC维,基于结构风险最小化理论对此问题进行分析,得出的结论跟实验结果相符. (err)