Multi-Label Classification
Multi-label classification is another type of classification where you predict not only one target variable as in binary or multi-class classification, but several response variables at the same time. For instance, you can predict multiple outputs for the different objects present in an image (for instance, a model will predict whether there is a cat, a man, and a car in a given picture) or you can predict multiple topics for an article (such as whether the article is about the economy, international news, and manufacturing).
Implementing a multi-label classification with neural networks is extremely easy, and you have already learned everything required to build one. In TensorFlow, a multi-label classifier's architecture will look the same as for multi-class, with a final output layer with multiple units corresponding to the number of target variables you want to predict. But instead of using softmax as the activation function and categorical...
