Deep learning

Includes bibliographical references and index.

Artificial Intelligence is a disruptive technology across business and society. There are three long-term trends driving this AI revolution: the emergence of Big Data, the creation of cheaper and more powerful computers, and development of better algorithms for processing an learning from data. Deep learning is the subfield of Artificial Intelligence that focuses on creating large neural network models that are capable of making accurate data driven decisions. Modern neural networks are the most powerful computational models we have for analyzing massive and complex datasets, and consequently deep learning is ideally suited to take advantage of the rapid growth in Big Data and computational power. In the last ten years, deep learning has become the fundamental technology in computer vision systems, speech recognition on mobile phones, information retrieval systems, machine translation, game AI, and self-driving cars. It is set to have a massive impact in healthcare, finance, and smart cities over the next years. This book is designed to give an accessible and concise, but also comprehensive, introduction to the field of Deep Learning. The book explains what deep learning is, how the field has developed, what deep learning can do, and also discusses how the field is likely to develop in the next 10 years. Along the way, the most important neural network architectures are described, including autoencoders, recurrent neural networks, long short-term memory networks, convolutional networks, and more recent developments such as Generative Adversarial Networks, transformer networks, and capsule networks. The book also covers the two more important algorithms for training a neural network, the gradient descent algorithm and Backpropagation