开发工具:
文件大小: 4kb
下载次数: 0
上传时间: 2012-04-18
详细说明： Matlab implementation for nonlinear robust fuzzy PCA algorithm Function call: [scores,loads]=nfrpca(x,LV,method) INPUTS: 1)x: your data matrix samples in rows and variables in columns. 2)LV: How many variables to use from data, if not specified all variables are used. 3)method: Optional, will be selected automatically. By writing 'nfrpca1' current method is selected. In current version only this one is implemented and will be selected automatically. OUTPUTS: scores: The principal component scores; that is, the repre sentation of X in the principal component space. Rows of SCORE correspond to observations, columns to components. loads: principal component coefficients also known as loadings. One can use this function in a following way: 1) load your data in your matlab For example in this case write load exampledata3.txt which should give you following data: exampledata3 = 0.4600 0.3400 0.1400 0.0300 0.5000 0.3400 0.1500 0.0200 0.4400 0.2900 0.1400 0.0200 0.7600 0.3000 0.6600 0.2100 0.4900 0.2500 0.4500 0.1700 0.7300 0.2900 0.6300 0.1800 there we have an artificial data where in four columns we have data measurements from six samples. 2) run the program by writing i.e. [scores,loads]=nfrpca(exampledata3,4,'nfrpca1') You will get scores and loads i.e. (depend on parameters you chooce) scores = -0.5171 -0.2720 -0.0730 -0.0308 -0.5499 -0.2841 -0.0733 -0.0067 -0.4866 -0.2394 -0.0602 -0.0019 -1.0710 0.0095 0.0141 -0.0023 -0.7283 -0.0001 -0.0002 -0.0602 -1.0228 0.0031 -0.0029 0.0099 loads = -0.7113 -0.4603 0.2561 0.4654 -0.2898 -0.4887 -0.5115 -0.6447 -0.6134 0.7343 -0.2857 -0.0541 -0.1837 0.1011 0.7689 -0.6040 Notice that in lines 40 to 43 you have parameters which you can vary to alter the results. Defaults now being: a=10; % parameter value in monotonic function y=tanh(a*x) ALFA0=1; % learning coefficient (0,1] ETA=.1; % soft threshold, a small positive number EXPO=1.5; % fuzziness variable ...展开收缩
(系统自动生成,下载前可以参看下载内容)
下载文件列表
相关说明
- 本站资源为会员上传分享交流与学习,如有侵犯您的权益,请联系我们删除.
- 本站是交换下载平台，提供交流渠道，下载内容来自于网络，除下载问题外，其它问题请自行百度。
- 本站已设置防盗链，请勿用迅雷、QQ旋风等多线程下载软件下载资源，下载后用WinRAR最新版进行解压.
- 如果您发现内容无法下载，请稍后再次尝试；或者到消费记录里找到下载记录反馈给我们.
- 下载后发现下载的内容跟说明不相乎，请到消费记录里找到下载记录反馈给我们，经确认后退回积分.
- 如下载前有疑问，可以通过点击"提供者"的名字，查看对方的联系方式，联系对方咨询.
