Geek Answers Handbook

I need to do SVM in weka to filter documents using Java

I am an absolute beginner. Never did a classifier or anything in weka using Java. I have used the interface before. Basically, I got a little lost. I looked at the filter class for weka and played a little with it. My documents are text documents, and I need to divide them into 2 categories.

I'm not sure how I define categories or how I upload documents to an IDE that will be classified

: - (

Any help / tutorials or pointers are appreciated.

+7
source share
2 answers

I found this java tutorial very useful, although very few resources are available on the Internet (which I found)

http://www.cs.waikato.ac.nz/ml/weka/index_documentation.html

hope this helps

+2
source

Using weka for the first time is a pain, but you will need to go through it.

In addition, I tried weka, but I had to reset it due to exceptions from the JVM from memory. I wrote my own small cluster algorithm using Ruby, the performance was much better.

Anyway, here's how to use SVM in WEKA:

  • You can follow this guide to using SVM in weka: www.stat.nctu.edu.tw/~misg/WekaInC.ppt

  • Now you need data in the ARFF format (and I recommend that you use it, it seems to me that this helps, the data looks more structured from WEKA). That way, you can do this using the XML2ARFF-Converter , which I wrote for myself. You can change it to read text files and convert the text file to ARFF.

+2
source

More articles:


All Articles