[Solved] Learner with numerical input and binominal output?
Dear all,
I want to build a series prediction process but I am not quite sure which learner is the right one to use.
The input is a multivariate numerical data set (including a numerical label).
The output is supposed to be a simple binominal "up" or "down".
I wanted to ask whether there is a special learner or setting for this kind of task. If not I could imagine to use a SVM with numerical output followed by a "generate attribute" operator that includes a formula like
if(prediction[for t+1]>=label[of t] then "up" else "down")
Best regards
Sachs
I want to build a series prediction process but I am not quite sure which learner is the right one to use.
The input is a multivariate numerical data set (including a numerical label).
The output is supposed to be a simple binominal "up" or "down".
I wanted to ask whether there is a special learner or setting for this kind of task. If not I could imagine to use a SVM with numerical output followed by a "generate attribute" operator that includes a formula like
if(prediction[for t+1]>=label[of t] then "up" else "down")
Best regards
Sachs