Depth learning to learn, I want to "Python depth study" should be required reading for most machine learning book lovers. The biggest advantage of the book is a framework that provides a "holistic view" to establish a complete map in the brain, which does not know what commonly used then accordingly targeted leak filled is more convenient, but if the Direct Access API documentation face a flood of often at a loss.
The book is divided into two parts, the first part for the introduction of the global depth study, including its relationship learning and artificial intelligence, machine, such as some of the basic concepts related tensor (Tensor), gradient descent, neural networks, back propagation algorithm and many more. Chapter three of which gave a simple example, correspond to the task of dichotomous, multi-classification and regression, allowing readers to quickly understand the basic use of Keras, familiar with the depth of a typical learning process the data flow problem. The second part is about the depth of learning practical applications in computer vision and natural language processing to highlight the convolution neural networks and recurrent neural network, told VAE and GAN.
These content from another perspective of the gradient something about loss, deep learning applications, and these applications Inspiration. It is also suitable for advanced.
"Python deep learning" HD Chinese PDF, 314 pages, catalogs and bookmarks, color with pictures, can be replicated; HD English PDF, 386 pages, catalogs and bookmarks, color with pictures, can be replicated; the English version of the two can be compared to learning . Complete source code.
Network disk download: http://106.13.73.98
Better title should be based Keras depth study. Keras is the author of developers, and therefore do not have the authority to say. Book biggest feature is not used a mathematical formula, and only code language to explain all aspects of the depth of learning (in addition to reinforcement learning), it is the best materials programmer population.