I am working on a decision making system, something about concert prices prediction to maximize the profit. Because it is multi-output, now data mining algorithm I know only neural network is suitable for multi-output, but the result is different every time. So I want to try SVM, however, SVM is only suit for single output, so I want to improve SVM with multi-output, but I don't know is it feasible? BTW, these outputs has relationship, I can't output these independently by running SVM several times.

  • $\begingroup$ possible duplicate of Best way to perform multiclass SVM $\endgroup$ – gung - Reinstate Monica Dec 25 '14 at 3:46
  • $\begingroup$ Actually, I use the SVR for regression, so it is not a multi-class problem, it is a multi-output regression problem. $\endgroup$ – Wendy LEE Dec 25 '14 at 3:52
  • $\begingroup$ In the sense of multiple different continuous response variables, or in the sense that the single continuous response variable can take multiple values? $\endgroup$ – gung - Reinstate Monica Dec 25 '14 at 3:55
  • $\begingroup$ For example, the concert tickets have several prices, this model is to predict the related price, so it has several prices. maybe 5 or 6 output $\endgroup$ – Wendy LEE Dec 25 '14 at 4:02
  • $\begingroup$ OK, maybe that's different enough. I retracted my close vote. We can see what the SVM experts think. $\endgroup$ – gung - Reinstate Monica Dec 25 '14 at 4:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.