TensorFlow is an open source software library for machine learning and training neural networks. TensorFlow was originally developed by Google, and was made open source in 2015.
Over the course of this book, you will learn how to use TensorFlow to solve a novel research problem. You'll use one of the most popular machine learning approaches, neural networks with TensorFlow. We'll work on both the simple and deep neural networks to improve our models.
You'll study images of letters and digits in various fonts with the goal of identifying fonts based on one specific image of a single letter. This will be a straightforward classification problem.
As no single pixel or position—but local structures among pixels—is important, it's an ideal problem for deep learning with TensorFlow. Though we'll start with simple models, this series will gradually introduce more nuanced approaches and explain the code line by line. By the end of this book, you'll have created your own advanced model for font recognition.
So let's put on our helmets; we're going deep into data mines with TensorFlow.
Chapter 1, Getting Started, discusses the techniques and the models we'll apply using TensorFlow. In this chapter, we will install TensorFlow on a machine we can use. After some small steps with basic computations, we will jump into a machine learning problem, successfully building a decent model with just logistic regression and a few lines of TensorFlow code.
Chapter 2, Deep Neural Networks, shows TensorFlow in its prime with deep neural networks. You will learn about the single and multiple hidden layer model. You will also learn about the different types of neural networks and build and train our first neural network with TensorFlow.
Chapter 3, Convolutional Neural Networks, talks about the most powerful developments in deep learning and applies the concepts of convolution to a simple example in TensorFlow. We will tackle the practical aspects of understanding convolution. We will explain what a convolutional and pooling layer is in a neural net, following with a TensorFlow example.
Chapter 4, Introducing Recurrent Neural Networks, introduces the concept of RNN models, and their implementation in TensorFlow. We will look at a simple interface to TensorFlow called TensorFlow learn. We will also walk through dense neural networks as well as understand convolutional neural networks and extracting weights in detail.
Chapter 5, Wrapping Up, wraps up our look at TensorFlow. We'll revisit our TensorFlow models for font classification, and review their accuracy.