guiji

说明：  ofdm同步技术的房展（对于学习ofdm同步技术有所把帮助）(ofdm synchronization technology房展(ofdm synchronization technology for learning has to help))

用于解决机器学习中分类问题，里面包含python，matlab和octave的接口及示例。(LIBLINEAR is a simple package for solving large-scale regularized linear classification.)

使用主成分分析对高维数据进行降维，程序包含了降维的过程和结果(Using PCA to reduct dimensionality for data. the program contains the process and solution.)

说明：  matlab将STK中卫星轨道数据导出函数(interface for transforming STK data into Matlab)

This is a GUI based system. The GUI is invoked from the command line by typing "ui_start". The GUI is self explanatory. Tool tips are given in blue color. The cursor needs to be positioned on these for detailed explanation. The main file is "runsim.m".It is recommended that 10000 runs (packets) be used for plotting frame error rates and 5000 runs for BER/SER. Obviously, you would require a fast machine!!! At least a P4 machine. We have used trace criterion for 4PSK states 4,8 and 16 as an example for 2x2 cases. The reader can use rank and determinant criteria, by commenting the trace criterion lines and copying the rank and determinant lines from the respective 2x1 cases. The SNR range can be varied by correcting the range at line 36 of file "runsim.m". This software has been tested with MATLAB version 6.0 and above with signal processing and communications toolbox options. (This is a GUI based system. The GUI is invoked from the command line by typing " ui_start" . The GUI is self explanatory. Tool tips are given in blue color. The cursor needs to be positioned on these for detailed explanation. The main file is " runsim.m" . It is recommended that 10000 runs (packets) be used for plotting frame error rates and 5000 runs for BER/SER. Obviously, you would require a fast machine!!! At least a P4 machine. We have used trace criterion for 4PSK states 4,8 and 16 as an example for 2x2 cases. The reader can use rank and determinant criteria, by commenting the trace criterion lines and copying the rank and determinant lines from the respective 2x1 cases. The SNR range can be varied by correcting the range at line 36 of file " runsim.m" . This software has been tested with MATLAB version 6.0 and above with signal processing and communications toolbox options.)

数据的预处理方法，极差归一化，用于对大量数据的预处理(Poor data pretreatment method, normalization, designed for large amounts of data pretreatment)

用MATLAB实现图像的直方图均衡化的处理，使图像更分明。(MATLAB image histogram equalization processing, make the image more clearly.)

信号预处理程序，包括最小二乘法，滑动平均法，五点三次法以及五点滑动法(Signal preprocessing programs, including least square, sliding average,five point three average and five-point sliding)

显示奇怪非混沌吸引子. 获得数据后, 用外部软件: origin等画出.(showed strange non-chaotic attractor. Access to data, the use of external software : origin such as paint.)

SVM神经网络的系统分类算法，修改可用，运行环境为Matlab。(System classification algorithm, SVM neural network to modify the available, operating environment for the Matlab.)