Machine Learning has revolutionized the tech industry for quite some time now. The newest addition to the ML Family which is taking the world by storm is Deep Learning- the extension of Artificial Neural Networks which has more than one hidden processing layers. These extra hidden layers have proved to be of phenomenal importance in the field of Computer Vision, Speech Recognition, Machine Translation, Natural Language Processing (NLP) , bioinformatics etc. where machines have produced an accuracy almost comparable to those of humans, and sometimes have resulted in super- human accuracy as well (i.e. more accurate than an average human mind).

Deep Learning has been the area of research for many for the past few years and three experts of the field- Ian Goodfellow, Yoshua Bengio and Aaron Courville, have come together for a masterpiece in the form of their book, ‘Deep Learning’.

The AI Bible in making

‘Deep Learning’ is basically organized into three main sections:

• Applied Math and Machine Learning Basics

• Modern Practical Deep Networks

• Deep Learning Research

The first section welcomes even the novice in the field, providing the mathematical background needed to fully understand the fundamentals of the concept. Although a pre- requisite knowledge of linear algebra concepts like Singular Value Decomposition (SVD) and Principal Component Analysis (PCA) , gradient based optimization techniques etc. This section creates a strong base for further topics and implementations.

The crux of the whole book lies in the second section about deep networks. It introduces the reader to advanced ML and ANN concepts like back- propagation, regularization parameter, hidden units, deep learning, and teach how to train deep models, CNNs (Convolutional Neural Networks), and RNNs (Recursive Neural Networks) for solving deep learning problems. The section is wrapped up by explaining deep learning deployment techniques, using GPUs and examples on Computer Vision, Speech Recognition, and NLP.

Section III is all about inspiring people to dig more into the concept and gives an insight into new techniques which are developing in the academic world, focusing more on the on- going research techniques into making a true AI system. This book can definitely be considered as a wholesome package for everyone who wishes to make a career in the field, be it in academia or in the industry.

About the authors

Yoshua Bengio is a full time Professor in the Department of Computer Science and Operations Research at Université de Montréal and Head at Montreal Institute for Learning Algorithms (MILA), and is also the co- director of the Learning in Machines and Brains project of the Canadian Institute of Advanced Research.

Aaron Courville is an Assistant Professor in the Department of Computer Science and Operations Research at Université de Montréal and member of the LISA lab (LISA: Laboratoire d’Informatique des Systèmes Adaptatifs). His current research interests are development of deep learning models and methods, and is inclined towards developing probabilistic models and novel inference methods.

Ian J. Goodfellow is a Staff Research Scientist working in the field of machine learning, and is currently employed as a research scientist at Google Brain. He has made several contributions to the field of deep learning

More information: DeepLearning Book