For binary classification issues, squared hinge loss is an alternative to cross-entropy that can be used as a loss function. While it can be used