How can I predict my boolean field?

I am trying to create a new sample with Peltarion. In my .csv file, I have 4 columns, first 3 are text fields, and last one is boolean.

I am trying to make a logic to predict my last boolean field (ie true or false from the input ie other 3 fields). But the status I am expecting is true/false.

I wonder why am getting numbers there. What should I do?

The value you are getting as status is the probability that the status is the positive class, which for you would be the status True.
A common way to proceed is to consider that if the probability is above 0.5 then it’s positive, and below 0.5 it’s negative. So 0.2513224 would mean False.

You can read more about choosing a value of threshold on the ROC curve page here.

1 Like