partition data set into training and test datasets

dhanya · Joined: 24 Jul 2011 Posts: 1

I have to divide a dataset into training(80%) and test datasets (20%)using random sampling. I have to do in c#.

Please help me how to do it....I have no idea...

alextui · Posted: Tue Oct 25, 2011 4:29 pm Post subject:

the job can NOT be easily done in C#, unless you had a very good framework.

You can try MATLAB or SAS

editor

You need to create an additional variable/column in your data, let's call it rnd which will have random values from 0 to 1.

To select 20% test data you assign to test data records where rnd<=0.2