There are a number of popular evaluation metrics for classification other than accuracy such as recall, precision, AUC, F-scores etc. Instead of listing them all here, I think it is best to point you towards some interesting resources that can kick-start your search for answers. Although you might not be using scikit, the metrics remain relevant. It also quite lists differences between binary classification and multi-class classification setting.
Sep-17-2018, 06:27:57 GMT