For me, the Experimenter as suggested by "Martin" is an excellent way.

Kind regards,

On Fri, Mar 25, 2016 at 11:56 PM, Martin <> wrote:
I believe the most reliable way to compare different classifiers in WEKA can be accomplished using WEKA "Experimenter".


On 25 March 2016 at 22:38, Francesca Rig [via WEKA] <[hidden email]> wrote:
I'm quite new to weka and data mining and I have to develop a project.

My idea is to use data from ended auctions of Ebay (Category of the item, shipping cost, auction duration, condition of the item, user's feedbacks) to build a model to predict the probability of sale of a given item.
The dataset that I have built is composed by those attributes:
end_price (numeric)
shipping_cost (numeric)
item_condition (nominal)
item_category (nominal)
auction_duration (nominal)
user_feedback (numeric)
user_feedback% (numeric)
end_auction_day (nominal)
and the class attribute is the "item_sold" one, with two possible values: YES or NO.

For my purpose I was looking for a classifier that outputs the probability that an item will belong to class YES, so that if a user inputs "item_category,auction_duration,feedback,shipping,cost" he will then receive as an output the probability of selling his item.

I have tried with bayesian classifiers but I am not sure they are the right choice to this aim.
Any hint on the way I should take to find the right classifier? Is this kind of classification problem feasible?

Thank you very much!

If you reply to this email, your message will be added to the discussion below:
To unsubscribe from WEKA, click here.

View this message in context: Re: Classifier to predict the probability of belonging to a class
Sent from the WEKA mailing list archive at

Wekalist mailing list
Send posts to:
List info and subscription status:
List etiquette: