Preface

Deep Learning Quick Reference demonstrates a fast and practical approach to using deep learning. It's focused on real-life problems, and it provides just enough theory and math to reinforce the readers' understanding of the topic. Deep learning is an exciting, fast paced branch of machine learning, but it's also a field that can be broken into. It's a field where a flood of detailed, complicated research is created every day, and this can be overwhelming. In this book, I focus on teaching you the skills to apply deep learning on a variety of practical problems. My greatest hope for this book is that it will provide you with the tools you need to use deep learning techniques to solve your machine learning problems.

Who this book is for

I'm a practicing data scientist, and I'm writing this book keeping other practicing data scientists and machine learning engineers in mind. If you're a software engineer applying deep learning, this book is also for you.

If you're a deep learning researcher, then this book isn't really for you; however, you should still pick up a copy so that you can criticize the lack of proofs and mathematical rigor in this book.

If you're an academic or educator, then this book is definitely for you. I've taught a survey source in data science at the University of Illinois at Springfield (go Prairie Stars!) for the past 3 years, and in doing so, I've had the opportunity to inspire a number of future machine learning people. This experience has inspired me to create this book. I think a book like this is a great way to help students build interest in a very complex topic.

What this book covers

Chapter 1, The Building Blocks of Deep Learning, reviews some basics around the operation of neural networks, touches on optimization algorithms, talks about model validation, and goes over setting up a development environment suitable for building deep neural networks.

Chapter 2, Using Deep Learning to Solve Regression Problems, enables you build very simple neural networks to solve regression problems and explore the impact of deeper more complex models on those problems.

Chapter 3, Monitoring Network Training Using TensorBoard, lets you get started right away with TensorBoard, which is a wonderful application for monitoring and debugging your future models.

Chapter 4, Using Deep Learning to Solve Binary Classification Problems, helps you solve binary classification problems using deep learning.

Chapter 5, Using Keras to Solve Multiclass Classification Problems, takes you to multiclass classification and explores the differences. It also talks about managing overfitting and the safest choices for doing so.

Chapter 6, Hyperparameter Optimization, shows two separate methods for model tuning—one, well-known and battle tested, while the other is a state-of-the-art method.

Chapter 7, Training a CNN From Scratch, teaches you how to use convolutional networks to do classification with images.

Chapter 8, Transfer Learning with Pretrained CNNs, describes how to apply transfer learning to get amazing performance from an image classifier, even with very little data.

Chapter 9, Training an RNN from scratch, discusses RNNs and LSTMS, and how to use them for time series forecasting problems.

Chapter 10, Training LSTMs with Word Embeddings From Scratch, continues our conversation on LSTMs, this time talking about natural language classification tasks.

Chapter 11, Training Seq2Seq Models, helps us use sequence to sequence models to do machine translation.

Chapter 12, Using Deep Reinforcement Learning, introduces deep reinforcement learning and builds a deep Q network that can power autonomous agents.

Chapter 13, Generative Adversarial Networks, explains how to use generative adversarial networks to generate convincing images.

To get the most out of this book

1. I assume that you're already experienced with more traditional data science and predictive modeling techniques such as Linear/Logistic Regression and Random Forest. If this is your first experience with machine learning, this may be a little difficult for you.
2. I also assume that you have at least some experience in programming with Python, or at least another programming language such as Java or C++.
3. Deep learning is computationally intensive, and some of the models we build here require an NVIDIA GPU to run in a reasonable amount of time. If you don't own a fast GPU, you may wish to use a GPU-based cloud instance on either Amazon Web Services or Google Cloud Platform.

Download the example code files

You can download the example code files for this book from your account at www.packtpub.com. If you purchased this book elsewhere, you can visit www.packtpub.com/support and register to have the files emailed directly to you.

You can download the code files by following these steps:

1. Log in or register at www.packtpub.com.
2. Select the SUPPORT tab.
3. Click on Code Downloads & Errata.
4. Enter the name of the book in the Search box and follow the onscreen instructions.

Once the file is downloaded, make sure that you unzip or extract the folder using the latest version of any of these:

• WinRAR/7-Zip for Windows
• Zipeg/iZip/UnRarX for macOS
• 7-Zip/PeaZip for Linux

The code bundle for the book is also hosted on GitHub at https://github.com/PacktPublishing/Deep-Learning-Quick-Reference. We also have other code bundles from our rich catalog of books and videos available at https://github.com/PacktPublishing/. Check them out!

Conventions used

There are a number of text conventions used throughout this book.

CodeInText: Indicates code words in text, database table names, folder names, filenames, file extensions, pathnames, dummy URLs, user input, and Twitter handles. Here is an example: "That's exactly what the ModelCheckpoint callback does for us."

A block of code is set as follows:

def binary_accuracy(y_true, y_pred):
    return K.mean(K.equal(y_true, K.round(y_pred)), axis=-1)

When we wish to draw your attention to a particular part of a code block, the relevant lines or items are set in bold:

def build_network(input_features=None):
    inputs = Input(shape=(input_features,), name="input")
    x = Dense(32, activation='relu', name="hidden1")(inputs)
    x = Dense(32, activation='relu', name="hidden2")(x)
    x = Dense(32, activation='relu', name="hidden3")(x)
    x = Dense(32, activation='relu', name="hidden4")(x)
    x = Dense(16, activation='relu', name="hidden5")(x)
    prediction = Dense(1, activation='linear', name="final")(x)
    model = Model(inputs=inputs, outputs=prediction)
    model.compile(optimizer='adam', loss='mean_absolute_error')
    return model

Any command-line input or output is written as follows:

model-weights.00-0.971304.hdf5
model-weights.02-0.977391.hdf5
model-weights.05-0.985217.hdf5

Bold: Indicates a new term, an important word, or words that you see onscreen. For example, words in menus or dialog boxes appear in the text like this. Here is an example: "Select System info from the Administration panel."

Warnings or important notes appear like this.

Tips and tricks appear like this.

Get in touch

Feedback from our readers is always welcome.

General feedback: Email [e-mail] and mention the book title in the subject of your message. If you have questions about any aspect of this book, please email us at [e-mail].

Errata: Although we have taken every care to ensure the accuracy of our content, mistakes do happen. If you have found a mistake in this book, we would be grateful if you would report this to us. Visit www.packtpub.com/submit-errata, select your book, click on the Errata Submission Form link, and enter the details.

Piracy: If you come across any illegal copies of our works in any form on the internet, we would be grateful if you would provide us with the location address or website name. Please contact us at [e-mail] with a link to the material.

If you are interested in becoming an author: If there is a topic that you have expertise in and you are interested in either writing or contributing to a book, please visit authors.packtpub.com.

Reviews

Please leave a review. Once you have read and used this book, why not leave a review on the site that you purchased it from? Potential readers can then see and use your unbiased opinion to make purchase decisions; also, we at Packt can understand what you think about our products, and our authors can see your feedback on their book. Thank you!

For more information about Packt, please visit packtpub.com.