Geek Answers Handbook

Function Selection in MATLAB

I have a text classification dataset, ready for use in MATLAB. Each document is a vector in this data set, and the dimension of this vector is extremely high. In these cases, people usually perform some functions on vectors, such as those that you actually find the WEKA toolkit. Is there something similar in MATLAB? if I can’t offer an algorithm for me to do this ...? thanks

+6
matlab classification
source share
3 answers

MATLAB (and its toolbar) includes a number of functions related to the choice of function:

You can also find examples demonstrating the use of real data sets:

In addition, there are third-party toolbars:

Otherwise, you can always call your favorite functions from WEKA directly from MATLAB, since it includes JVM ...

+12
source share

The choice of function depends on the specific task that you want to do for text data.

One of the simplest and roughest methods is to use basic component analysis (PCA) to reduce data sizes. These reduced measurement data can be used directly as functions for classification.

See PCA usage guide here:

http://matlabdatamining.blogspot.com/2010/02/principal-components-analysis.html

Here is a link to help the Matlab PCA team:

http://www.mathworks.com/help/toolbox/stats/princomp.html

Using the obtained functions, well-known vector vector machines (SVM) can be used for classification.

http://www.mathworks.com/help/toolbox/bioinfo/ref/svmclassify.html http://www.autonlab.org/tutorials/svm.html

+1
source share

You can use the Weiss and Kulikovsky technique of independent functions to quickly eliminate variables that obviously don't matter:

http://matlabdatamining.blogspot.com/2006/12/feature-selection-phase-1-eliminate.html

+1
source share

More articles:


All Articles