▶Book Description
Deep learning enables efficient and accurate learning from a massive amount of data. This book will help you overcome a number of challenges using various deep learning algorithms and architectures with R programming.
This book starts with a brief overview of machine learning and deep learning and how to build your first neural network. You’ll understand the architecture of various deep learning algorithms and their applicable fields, learn how to build deep learning models, optimize hyperparameters, and evaluate model performance. Various deep learning applications in image processing, natural language processing (NLP), recommendation systems, and predictive analytics will also be covered. Later chapters will show you how to tackle recognition problems such as image recognition and signal detection, programmatically summarize documents, conduct topic modeling, and forecast stock market prices. Toward the end of the book, you will learn the common applications of GANs and how to build a face generation model using them. Finally, you’ll get to grips with using reinforcement learning and deep reinforcement learning to solve various real-world problems.
By the end of this deep learning book, you will be able to build and deploy your own deep learning applications using appropriate frameworks and algorithms.
▶What You Will Learn
- Design a feedforward neural network to see how the activation function computes an output
- Create an image recognition model using convolutional neural networks (CNNs)
- Prepare data, decide hidden layers and neurons and train your model with the backpropagation algorithm
- Apply text cleaning techniques to remove uninformative text using NLP
- Build, train, and evaluate a GAN model for face generation
- Understand the concept and implementation of reinforcement learning in R
▶Key Features
- Understand deep learning algorithms and architectures using R and determine which algorithm is best suited for a specific problem
- Improve models using parameter tuning, feature engineering, and ensembling
- Apply advanced neural network models such as deep autoencoders and generative adversarial networks (GANs) across different domains
▶Who This Book Is For
This book is for data scientists, machine learning engineers, and deep learning developers who are familiar with machine learning and are looking to enhance their knowledge of deep learning using practical examples. Anyone interested in increasing the efficiency of their machine learning applications and exploring various options in R will also find this book useful. Basic knowledge of machine learning techniques and working knowledge of the R programming language is expected.
▶What this book covers
- Chapter 1, Machine Learning Basics, reviews all the essential elements of machine learning. This quick refresher is important as we move into deep learning, a subset of machine learning, which shares a number of common terms and methods.
- Chapter 2, Setting Up R for Deep Learning, summarizes the common frameworks and algorithms for deep learning and reinforced deep learning in R. You will become familiar with the common libraries, including MXNet, H2O, and Keras, and learn how to install each library in R.
- Chapter 3, Artificial Neural Networks, teaches you about artificial neural networks, which make up the base building block for all deep learning. You will build a simple artificial neural network and learn how all of its components combine to solve complex problems.
- Chapter 4, CNNs for Image Recognition, demonstrates how to use convolutional neural networks for image recognition. We will briefly cover why these deep learning networks are superior to shallow nets. The remainder of the chapter will cover the components of a convolutional neural network with considerations for making the most appropriate choice.
- Chapter 5, Multilayer Perceptron Neural Networks for Signal Detection, shows how to build a multilayer perceptron neural network for signal detection. You will learn the architecture of multilayer perceptron neural networks, and also learn how to prepare data, define hidden layers and neurons, and train a model using a backpropagation algorithm in R.
- Chapter 6, Neural Collaborative Filtering Using Embeddings, explains how to build a neural collaborative filtering recommender system using layered embeddings. You will learn how to use the custom Keras API, construct an architecture with user-item embedding layers, and train a practical recommender system using implicit ratings.
- Chapter 7, Deep Learning for Natural Language Processing, explains how to create document summaries. The chapter begins with removing parts of documents that should not be considered and tokenizing the remaining text. Afterward, embeddings are applied and clusters are created. These clusters are then used to make document summaries. We will also learn to code a Restricted Boltzmann Machine (RBM) along with defining Gibbs Sampling, Contrastive Divergence, and Free Energy for the algorithm. The chapter will conclude with compiling multiple RBMs to create a deep belief network.
- Chapter 8, Long Short-Term Memory Networks for Stock Forecasting, shows how to use long short-term memory (LSTM) RNN networks for predictive analytics. You will learn how to prepare sequence data for LSTM and how to build a predictive model with LSTM.
- Chapter 9, Generative Adversarial Networks for Faces, describes the main components and applications of generative adversarial networks (GANs). You will learn the common applications of generative adversarial networks and how to build a face generation model with GANs.
- Chapter 10, Reinforcement Learning for Gaming, demonstrates the reinforcement learning method on a tic-tac-toe game. You will learn the concept and implementation of reinforcement learning in a highly customizable framework. Moreover, you will also learn how to create an agent that plays the best action for each game step and how to implement reinforcement learning in R.
- Chapter 11, Deep Q-Learning for Maze Solving, shows us how to use R to implement reinforcement learning techniques within a maze environment. In particular, we will create an agent to solve a maze by training an agent to perform actions and to learn from failed attempts.
