I am currently studying binary classification modeling, and have observed that two target classes are typically used to encode target labels: 0/1 or -1/1. Why are there two encoding classes and what are the pros and cons of each?

  • 1
    $\begingroup$ check my question above. People from statistics community like to encode with 0 and 1. and people from machine learning community like to code with 1 and -1. $\endgroup$ – Haitao Du Oct 4 '17 at 18:27