Deep Learning Quick Reference

所需积分/C币:10 2019-02-06 13:09:05 5.32MB PDF
收藏 收藏

Deep Learning Quick Reference Copyright o 2018 Packt Publishing All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without the prior written permission of the publisher, except in the case of brief quotations embedded in critical articles or reviews Every effort has been made in the preparation of this book to ensure the accuracy of the information presented However the information contained in this book is sold without warranty, either express or implied Neither the author, nor Packt Publishing or its dealers and distributors, will be held liable for any damages caused or alleged to have been caused directly or indirectly by this book Packt Publishing has endeavored to provide trademark information about all of the companies and products mentioned in this book by the appropriate use of capitals. However, Packt Publishing cannot guarantee the accuracy of this information Commissioning Editor: Amey Varangaonkar Acquisition editor: Virai madhav Content Development Editor: Varun Sony Technical Editor: Dharmendra yadav Copy Editors: Safis Editing Project Coordinator: Manthan Patel Proofreader: Safis editing Indexer: Pratik shirodkar Graphics: Tania dutta Production Coordinator: Deepika Naik First published: March 2018 Production reference: 1070318 Published by packt Publishing ltd Livery place 35 Livery street Birmingham B3 2PB, UK ISBN978-1-78883-799-6 To my wife Lana, whose love and support define the best epoch of my life To my son, William, who is likely disappointed that this book doesn t have more dragons in t To my mother, Sharon, and to the memory of my father, Bob, who taught me that determination and resilience matter more than intelligence Mapt mapt.lo Mapt is an online digital library that gives you full access to over 5,000 books and videos, as well as industry leading tools to help you plan your personal development and advance your career. For more information, please visit our website Why subscribe? Spend less time learning and more time coding with practical eBooks and Videos from over 4,000 industry professionals Improve your learning with skill plans built especially for you Get a free eBook or video every month mapt is fully searchable Copy and paste, print, and bookmark content PacktPub, com Did you know that packt offers e Book versions of every book published, with PDF and epubfilesavailableYoucanupgradetotheeboOkversionatwww.PacktPub.comandasa print book customer, you are entitled to a discount on the e Book copy. Get in touch with us at service@packtpub com for more details Atwww.packtpub.comyoucanalsoreadacollectionoffreetechnicalarticlessignupfora range of free newsletters and receive exclusive discounts and offers on packt books and eBooKs Foreword I first met Mike Bernico when we were two of the founding members of a new data science team at a fortune 50 company Then it was a head y time there wasn ' t such a thing as formal data science education, so we were all self-taught. We were a collection of adventurous people from diverse backgrounds, who identified and learned data science techniques because we needed them to solve the problems that we were interested in. We built a team with an optimistic hacker approach-the belief that we could find and apply techniques"from the wild"to build interesting, useful things It is in this practical, scrappy spirit that Mike wrote Deep learning Quick reference book Deep learning is frequently made out to be mysterious and difficult; however, in this guide, Mike breaks down major deep learning techniques, making them approachable and applicable. With this book, you (yes, you! )can quickly get started with using deep learning for your own projects in a variety of different modalities Mike has been practising data science since before the discipline was named, and he has been specifically teaching the topic to university students for 3 years. Prior to this, he spent many years as a working computer scientist with a specialization in networks and security and he also has a knack for engaging with people and communicating with nonspecialists He is currently the lead data scientist for a large financial services company where he designs systems for data science, builds machine learning models with direct applications and for research publications, mentors junior data scientists, and teaches stakeholders about data science. he knows his stuff With Deep learning quick reference book, you'll benefit from Mike's deep experience, humor, and down-to-earth manner as you build example networks alongside him. After you complete Mike's book, you'l have the confidence and knowledge to understand and apply deep learning to the problems of your own devising for both fun and function Bon voyage, and good hacking! J Malia Andrus, Ph. D Data Scientist Seattle washington Contributors About the author Mike bernico is a Lead data Scientist at State Farm Mutual Insurance companies he also works as an adjunct for the University of illinois at Springfield, where he teaches Essentials of Data Science, and Advanced Neural Networks and Deep Learning. Mike earned his MSCS from the University of Illinois at Springfield. He's an advocate for open source software and the good it can bring to the world. As a lifelong learner with umpteen hobbies, Mike also enjoys cycling, travel photography, and wine making Id like to thank the very talented State Farm Data Scientists, current and past, for their friendship expertise, and encouragement Thanks to my technical reviewers for providing insight and assistance with this book Most importantly, I'd like to thank my wife, Lana, and my son, Will, for making time for this in our lives About the reviewer Vitor Bianchi Lanzetta has a master's degree in Applied Economics from the University of Sao Paulo, one of the most reputable universities in Latin America he has done a lot of research in economics using neural networks. He has also authored r Data Visualization Recipes, Packt Publishing. Vitor is very passionate about data science in general, and he walks the earth with a personal belief that he is just as cool as he is geek. He thinks that you will learn a lot from this book, and that Tensor Flow may be the greatest deep learning tool currently available Packt is searching for authors like you If you re interested in becoming an author for Packt, please visit authors. packtpub com and apply today. We have worked with thousands of developers and tech professionals, just like you, to help them share their insight with the global tech community. You can make a general application, apply for a specific hot topic that we are recruiting an author for, or submit your own idea Table of contents Preface Chapter 1: The Building Blocks of Deep Learning The deep neural network architectures Neurons The neuron linear function 67777 Neuron activation functions The loss and cost functions in deep learning The forward propagation process The back propagation function 12 Stochastic and minibatch gradient descents 12 Optimization algorithms for deep learning 13 Using momentum with gradient descent 13 The RMsProp algorithm 14 The Adam optimizer 14 Deep learning frameworks 15 What is TensorFlow? 15 What is Keras? 16 Popular alternatives to TensorFlow 16 GPU requirements for Tensor Flow and Keras 16 Installing nvidia cuda toolkit and cudNN Installing Python 19 Installing tensor Flow and Keras 20 Building datasets for deep learning 22 Bias and variance errors in deep learning 22 The train val and test datasets 23 Managing bias and variance in deep neural networks 24 K-Fold cross-validation 24 Summary 25 Chapter 2: Using Deep Learning to Solve Regression Problems 26 Regression analysis and deep neural networks 26 Benefits of using a neural network for regression 27 Table of contents Drawbacks to consider when using a neural network for regression 28 Using deep neural networks for regression 28 How to plan a machine learning problem 28 Defining our example problem 29 Loading the dataset 29 Defining our cost function Building an MLP in Keras 31 Input layer shape 32 Hidden layer shape 32 Output layer shape 32 Neural network architecture 33 Training the Keras model 34 Measuring the performance of our model 35 Building a deep neural network in Keras 36 Measuring the deep neural network performance 37 Tuning the model hyperparameters 38 Saving and loading a trained Keras model 39 Summary 40 Chapter 3: Monitoring Network Training Using TensorBoard a brief overview of tensor board 41 Setting up TensorBoard 42 Installing tensor Board 42 How Tensor board talks to Keras/TensorFlow 43 Running Tensor Board 43 Connecting Keras to TensorBoard 44 Introducing Keras callbacks 44 Creating a Tensor Board callback 45 Using Tensor Board 48 Visualizing training 48 Visualizing network graphs 49 Visualizing a broken network Summary 52 Chapter 4: Using Deep Learning to Solve Binary Classification Problems 53 Binary classification and deep neural networks 54 [i]

试读 127P Deep Learning Quick Reference
立即下载 低至0.43元/次 身份认证VIP会员低至7折
  • 分享宗师

关注 私信
Deep Learning Quick Reference 10积分/C币 立即下载
Deep Learning Quick Reference第1页
Deep Learning Quick Reference第2页
Deep Learning Quick Reference第3页
Deep Learning Quick Reference第4页
Deep Learning Quick Reference第5页
Deep Learning Quick Reference第6页
Deep Learning Quick Reference第7页
Deep Learning Quick Reference第8页
Deep Learning Quick Reference第9页
Deep Learning Quick Reference第10页
Deep Learning Quick Reference第11页
Deep Learning Quick Reference第12页
Deep Learning Quick Reference第13页
Deep Learning Quick Reference第14页
Deep Learning Quick Reference第15页
Deep Learning Quick Reference第16页
Deep Learning Quick Reference第17页
Deep Learning Quick Reference第18页
Deep Learning Quick Reference第19页
Deep Learning Quick Reference第20页

试读结束, 可继续阅读

10积分/C币 立即下载 >