I am new and I do not have experience about Maxent theory.
I am using Maxent as Classifyer and I have 3 different outcomes (1,2,3). My sample data is made by:
- 160.000 records with outcome=1
- 26.688 records with outcome=2
- 4.614 records with outcome=3
so there is a big difference in the numebers of each type of outcome.
How much are influenced the statistics for the maxent model by this difference?
Is it correct to train the system with so different data or it should be better to reduce the nr. of records with outcome =1 and 2 in order to have similar nr. of records for each outcome?
thank you.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I would like to know if you are using net beans to run the experiments . I am having difficulty running the code itself. Wold really like to have some feedback.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
It seems you were able to run the experiments using some sample data to get the output of the classifier.
I would like to find out how can I run the experiments using sample data to get some output for a text file.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I am new and I do not have experience about Maxent theory.
I am using Maxent as Classifyer and I have 3 different outcomes (1,2,3). My sample data is made by:
- 160.000 records with outcome=1
- 26.688 records with outcome=2
- 4.614 records with outcome=3
so there is a big difference in the numebers of each type of outcome.
How much are influenced the statistics for the maxent model by this difference?
Is it correct to train the system with so different data or it should be better to reduce the nr. of records with outcome =1 and 2 in order to have similar nr. of records for each outcome?
thank you.
I would like to know if you are using net beans to run the experiments . I am having difficulty running the code itself. Wold really like to have some feedback.
I have used a model of Maxent made in C-sharp.
This is the link of what I used.
http://www.codeproject.com/KB/cs/sharpentropy.aspx?print=true
Thanks.
It seems you were able to run the experiments using some sample data to get the output of the classifier.
I would like to find out how can I run the experiments using sample data to get some output for a text file.