Hi,
I am trying to understand LUPI-SVM and I couldn't get much into the mathematical part. Can you help me in understanding the output of LUPI-SVM? If the labels are "0" and "1", I would expect a traditional classifier to output "0" or "1" or the prediction probability in range of 0 to 1. But while executing LUPI-SVM example, some of the predicted values are negative and some are greater than 1. Can you help me to understand how to get the predicted label or the probability from these values? Also, is there any assumptions to be considered with the features used for modelling?
Hi,
I am trying to understand LUPI-SVM and I couldn't get much into the mathematical part. Can you help me in understanding the output of LUPI-SVM? If the labels are "0" and "1", I would expect a traditional classifier to output "0" or "1" or the prediction probability in range of 0 to 1. But while executing LUPI-SVM example, some of the predicted values are negative and some are greater than 1. Can you help me to understand how to get the predicted label or the probability from these values? Also, is there any assumptions to be considered with the features used for modelling?