Unsupervised learning

Unsupervised learning is a form of learning in which there is no label provided to the learning algorithm along with the inputs, unlike the supervised learning paradigm. This class of learning algorithm is typically used to figure out patterns in the input data and cluster similar data together. A recent advancement in the field of deep learning introduced a new form of learning called Generative Adversarial Networks, which have gained massive popularity during the time this book was being written. If you are interested, you can learn a lot more about Generative Adversarial Networks from this video: https://www.packtpub.com/big-data-and-business-intelligence/learning-generative-adversarial-networks-video.