Thursday, February 10, 2011

RapidMiner : A how-to guide

In my business intelligence class we have been using a program called RapidMiner, which is an open-source system for data mining. In this class, we are analyzing almost 23 gigs of raw data, and it can seem very confusing at first. Unless you are running a very powerful computer, Rapid Miner can often freeze and crash due to the large amount of data and processing usage.

In this post, I will detail how to make RapidMiner work with your computer to its full potential, as well as detail how to use Rapid Miner.

Speeding Rapid Miner up:

1) Close all process you DONT need with task manager. Use common sense with this. Keep core processes and programs running but end everything unnecessary. (Example: Close things like limewire.exe or itunes.exe, but keep things like windows.exe and rapid-i.exe open!)

2) Find out how much RAM your system has. You can accomplish this by right clicking "My Computer" and clicking properties.

3) Go into your C: drive and then the program files folder. After that, find Rapid-I folder and open it up. Then,
a) for windows open up the rapidminerGUI
b) for Mac or linux, the other ones.

It should bring up a script in notepad.

Go down to where you see the line about JAVA MEMORY Max. Change it to something compatible for the amount of RAM your system has (remember to leave 1.5 - 2gb of ram for your OS.) For example, if you have a 2gb laptop, changing the amount to 1024 should be fine, although not preferable. It is a very powerful program and needs adequate resources to run.

This should get your RapidMiner working at full power.

Using RapidMiner:

Following are some video tutorials on how to use Rapid Miner. Good luck!

No comments:

Post a Comment