July 2019
Beginner to intermediate
740 pages
16h 52m
English
We evaluate the performance of classification models by looking at how well each class in the data was predicted by the model. The positive class is the class of interest to us; all other classes are considered negative classes. In our red wine classification, the positive class is high quality, while the negative class is low quality. Despite our problem only being a binary classification problem, the metrics that are discussed in this section extend to multiclass classification problems.
Read now
Unlock full access