Deep Time Series Forecasting with Python.pdf

【实例简介】
Deep Time Series Forecasting with Python. 使用深度学习技术进行时间序列回归预测
DEEP TIME SERIES FORECASTING With PYThon An Intuitive Introduction to Deep learn ing for applied Time series Modeling Dr, n.d lewis Copyright o 2016 by N D. Lewis All rights reserved. No part of this publication may be reproduced, dis- tributed, or transmitted in any form or by any means, including photo- copying, recording, or other electronic or mechanical methods, without the prior written permission of the author, except in the case of brief quo tations embodied in critical reviews and certain other noncommercial uses permitted by copyright law. For permission requests, contact the author t:www.Auscov.com Disclaimer: Although the author and publisher have made every effort to ensure that the information in this book was correct at press time, the author and publisher do not assume and hereby disclaim any liability to any party for any loss, damage, or disruption caused by errors or omissions whether such errors or omissions result from negligence, accident, or any other cause Ordering Information: Quantity sales. Special discounts are available on quantity purchases by corporations, associations, and others. For details emailinfo@nigeldlewis.cOm Image photography by Deanna Lewis with helpful assistance from Naomi lewis ISBN-13:978-1540809087 ISBN-10:1540809080 Contents Acknowledgements Preface vIll How to get the absolute most possible benefit from this book Getting Python Learning Python Using pack 13345 Additional resources to Check Out 1 The Characteristics of Time Series Data Simplified Understanding the Data Generating Mechanism Generating a Simple Time Series using Python Randomness and Reproducibility 12 The Importance of Temporal Order The Ultimate goal For Additional Exploration 15 2 Deep Neural Networks Explained 17 What is a Neural Network? The role of neuron Deep Learning in a Nutshell Generating Data for use with a Deep Neural Netwo Exploring the Sample Data 22 Translating Sample Data into a Suitable Format 25 A Super Easy Deep Neural Network Tool 26 Assessing model performance 2 Additional resources to Check Out 30 3 Deep Neural Networks for Time Series Forecasting the Easy Way 31 etting the data from the i 31 Cleaning up Downloaded Spreadsheet Files Understanding Activation Functions 36 How te 39 ssessln ng Partial Autocorrelation 42 a Neural Network Architecture for Time series Fo 45 Additional resources to Check Out 19 4 A Simple Way to Incorporate Additional Attributes in Your Model 51 Working with Additional Attributes 51 The Working of the Neuron Simplified How a Neural network learns Gradient Descent Clarified How to Easily Specify a Model 59 Choosing a Learning Rate The efficient Way to Run Your Model 62 Additional Resources to Check Out 66 5 The Simple Recurrent Neural Network 67 Why Use Keras? What is a Recurrent Neural Network? Gain Clarity on the role of the Delay Units 71 Follow this Approach to Create Your Train and Test Sets Parameter Sharing Clarified Understand Backpropagation Through Time 73 A Complete Intuitive guide to momentum 76 How to Benefit from Mini Batching 78 Additional Resources to Check Out 81 6 Elman neural networks 83 Prepare You Data for Easy Use 84 How to Model a Complex Mathematical Relationship with No Knowledge Use this Python library for Rapid Results Exploring the Error Surface A Super Simple Way to Fit the Model 91 Additional resources to Check Out 7 Jordan Neural networks 95 The Fastest Path to Data Preparation A Straightforward Module for Jordan Neural Networks Assessing Model Fit and Performance Additional Resources to Check Out .100 8 Nonlinear Auto-regressive Network with Exogenous Inputs 103 What is a narX Network? 103 Spreadsheet Files made easy with Panda 105 Working with macroeconomic Variables 107 Python and Pandas Data Types a Tool for Rapid narX Model Construction How to run the model 115 Additional Resources to Check Out 9 Long Short-Term Memory Recurrent Neural Network 119 Cyclical Patterns in Time Series Data 119 What is an LSTM? 122 Efficiently Explore and Quickly Understand Data 123 The LSTM Memory Block in a Nutshell 127 Straightforward Data Transformation for the Train and Test Sets .128 Clarify the role of Gates 130 Understand the Constant Error Carousel 131 Specifying a LstM Model the Easy Way 132 Shuffling Examples to Improve Generalization 136 A Note on Vanishing gradients Follow these Steps to build a Stateful LSTM ..139 Additional Resources to check Out 144 10 Gated recurrent unit 145 The Gated Recurrent Unit in a nutshell 145 A Simple approach to gated recurrent Unit Construction 148 a Quick Recap 150 How to Use Multiple Time Steps 151 Additional resources to Check Out ..154 11 Forecasting Multiple Outputs 155 Working with Zipped Files 156 How to Work with Multiple Targets .159 Creation of hand Crafted Feature 161 Model specification and Fit Additional Resources to Check Out 12 Strategies to Build Superior Models 169 Revisiting the UK Unemployment Rate Economic Data 169 Limitations of the Sigmoid Activation Function .171 One Activation Function You Need to Add to Your Deep Learning Toolkit.. 172 Try This Simple Idea to Enhance Success A Simple Plan for Early Stopping 180 Additional resources to check Out nex Dedicated to Angela, wife, friend and mother extraordinaire 【实例截图】
【核心代码】