kftool_matlab

说明：  内含8个数学建模的实验报告：线性规划、插值、拟合、非线性规划、蒙特卡洛方法、微分方程、无约束优化、回归分析(Mathematical modeling includes 8 Experiment Report: linear programming, interpolation, fitting, linear programming, Monte Carlo method, differential equations, unconstrained optimization, regression analysis)

说明：  用格拉布斯准则判断异常数据，通过检测输入数据剔除粗大误差( use Grubbs Criteria to determine outlier ， removing the gross error detection )

用于图像分块，将图像分为大小一致的图像，便于图像像素位置的确定(For the image block, the image into an image of the same size, easy to determine the location of the image pixel)

小波包分析提取振动信号中的特征频率，阐述了负荷预测的应用研究,matlab编写的元胞自动机。( Wavelet packet analysis to extract vibration signal characteristic frequency, It describes the application of load forecasting, matlab prepared cellular automata.)

DNN程序，可以利用MATLAB实现深度神经网络。包含子程序(DNN program, you can use MATLAB to achieve a deep neural network. Include subroutine)

一篇关于多相滤波器的论文,讲解了关于多信道的实现与仿真。(A paper about the polyphase filter，explained about the realization of multi-channel and simulation )

说明：  在matlab环境下的一个朴素贝叶斯分类器，大家可以看下(Matlab environment a Naive Bayes classifier, we can look)

用MATLAB实现有限长滤波器的源代码，东南大学第二次实验(MATLAB realization of finite length filter with the source code, Southeast University, the second experiment)

基于逆系统法的荷载识别程序，对于模型简单的识别效果较好(Based on the method of inverse system load identification procedures for the simple model, good recognition effect)

本软件主要用于结构的时域试验模态分析，包括单输入单输出，单输入多输出和多输入多输出程序设计。传统的模态分析软件大多是基于频域分析方法，需首先对采样信号进行Fourier 变换。本软件使用实测时域信号，无需Fourier变换，可以避免由于信号截断而引起泄露、出现旁瓣、分辨率低等问题对辨识精度的不利影响。此外，节省了计算机运算时间，可以对实际工况状态下的结构进行在线参数识别。(This software is mainly used for the structure of time-domain experimental modal analysis, including single-input single-output, single-input multiple-output and multiple-input multiple-output programming. Traditional modal analysis software are mostly based on frequency domain analysis, needed first sampled signal Fourier transform. The software uses the measured time-domain signal without the need for Fourier transform, avoid truncated because of the signal caused by leaks, there sidelobes and low resolution of the issue of adverse effects on the identification accuracy. In addition, the savings in computing time, to the actual condition of the structure of the state-line parameter identification.)