Recurrent Neural Networks with Python Quick Start Guide
上QQ阅读APP看书,第一时间看更新

Preface

Deep learning (DL) is an increasingly popular topic that attracts the attention of the largest corporations as well as that of all kinds of developers. Over the past five years, this field has seen massive improvements that have ultimately led us to think of DL as a highly disruptive technology with immense potential. Virtual assistants, speech recognition, and language translation are just a few examples of the direct implementation of DL techniques. Compared to image recognition or object detection, these applications use sequential data, where the nature of every result depends upon that of the previous one. For example, you can't produce a meaningful translation of a sentence from English to Spanish without tracking the words from beginning to end. For these kinds of problems, a specific type of model is being used—the recurrent neural network (RNN). In this book, we will cover the basics of RNNs and focus on some practical implementations using the popular DL library TensorFlow. All examples are accompanied by in-depth explanations of the theory to help you understand the underlying concepts behind this powerful but slightly complex model. Reading this book will leave you confident in your knowledge of RNNs and give you a good head start in using this model for your own specific use cases.