Deep Learning for Time Series Cookbook: Use PyTorch and Python recipes for forecasting, classification, and anomaly detection - Softcover

Vitor Cerqueira; Luís Roque

Softcover

ISBN 10: 1805129236 ISBN 13: 9781805129233

Publisher: Packt Publishing, 2024

Synopsis

Build real-world Artificial Intelligence applications with Time Series data and Deep Learning

Take your deep learning skills to the next level by mastering PyTorch with tens of Python recipes

Solve forecasting problems and predict the future using advanced neural network architectures in PyTorch

Key Features

• Learn how to train accurate forecasting model using neural networks and real-world time series
• Build advanced deep neural network architectures using PyTorch
• Tackle several time series tasks, such as forecasting, classification, hierarchical forecasting, and anomaly detection

Book Description

Many real-world systems are captured through the lens of time series. The analysis and forecasting of time series has thus become a key aspect of several organizations. Deep learning is the hottest Artificial Intelligence technology. It leverages large amounts of data to build intricate and accurate forecasting models.

This book is a comprehensive cookbook that guides you through the development of deep learning models for time series data using PyTorch. We start from the basic concepts behind time series analysis and the PyTorch framework. Then, we dive into the details of several time series problems, including forecasting, classification, anomaly detection, and hierarchical time series forecasting. You'll learn how to tackle these tasks with a set of code recipes.

By the end of this book, you'll have a solid understanding of time series data problems and how to tackle them using deep learning based on PyTorch.

What you will learn

• Understand main time series analysis concepts and how to apply them using pandas
• Learn about PyTorch and how to use it to build deep learning models
• Explore how to transform a time series for training transformers and other advanced deep neural networks
• Understand how to deal with various time series characteristics, such as trend, seasonality, or non-constant variance
• Tackle different kinds of forecasting problems, involving univariate, multivariate, or hierarchical time series
• Understand how to apply residual and convolutional neural networks for time series classification problems
• Learn how to solve time series anomaly detection problems using auto-encoders and Generative Adversarial Networks

Who This Book Is For

If you are a machine learning enthusiast or someone who wants to learn more about building forecasting applications using deep learning, this book is for you. In order to learn from this book, you should have basic knowledge of Python and machine learning.

Table of Contents

1. Getting Started with Time Series
2. Getting Started with keras
3. Univariate Time Series Forecasting
4. Advanced Forecasting Problems
5. Advanced Deep Learning Architectures for Time Series Forecasting
6. Probabilistic Time Series Forecasting
7. Deep Learning for Time Series Classification
8. Deep Learning for Time Series Anomaly Detection

About this title

About the Authors

​Vitor Cerqueira is a time series researcher with an extensive background in machine learning. Vitor obtained his Ph.D. degree in Software Engineering from the University of Porto in 2019. He is currently a Post-Doctoral researcher in Dalhousie University, Halifax, developing machine learning methods for time series forecasting. Vitor has co-authored several scientific articles that have been published in multiple high-impact research venues.

Luís Roque, is the Founder and Partner of ZAAI, a company focused on AI product development, consultancy, and investment in AI startups. He also serves as the Vice President of Data & AI at Marley Spoon, leading teams across data science, data analytics, data product, data engineering, machine learning operations, and platforms. In addition, he holds the position of AI Advisor at CableLabs, where he contributes to integrating the broadband industry with AI technologies. Luís is also a Ph.D. Researcher in AI at the University of Porto's AI&CS lab and oversees the Data Science Master's program at Nuclio Digital School in Barcelona. Previously, he co-founded HUUB, where he served as CEO until its acquisition by Maersk.

From the Back Cover

Learn how to deal with time series data and how to model it using deep learning and take your skills to the next level by mastering PyTorch using different Python recipesKey FeaturesLearn the fundamentals of time series analysis and how to model time series data using deep learning Explore the world of deep learning with PyTorch and build advanced deep neural networks Gain expertise in tackling time series problems, from forecasting future trends to classifying patterns and anomaly detection Purchase of the print or Kindle book includes a free PDF eBook Book Description Most organizations exhibit a time-dependent structure in their processes, including fields such as finance. By leveraging time series analysis and forecasting, these organizations can make informed decisions and optimize their performance. Accurate forecasts help reduce uncertainty and enable better planning of operations. Unlike traditional approaches to forecasting, deep learning can process large amounts of data and help derive complex patterns. Despite its increasing relevance, getting the most out of deep learning requires significant technical expertise. This book guides you through applying deep learning to time series data with the help of easy-to-follow code recipes. You'll cover time series problems, such as forecasting, anomaly detection, and classification. This deep learning book will also show you how to solve these problems using different deep neural network architectures, including convolutional neural networks (CNNs) or transformers. As you progress, you'll use PyTorch, a popular deep learning framework based on Python to build production-ready prediction solutions. By the end of this book, you'll have learned how to solve different time series tasks with deep learning using the PyTorch ecosystem.What you will learnGrasp the core of time series analysis and unleash its power using Python Understand PyTorch and how to use it to build deep learning models Discover how to transform a time series for training transformers Understand how to deal with various time series characteristics Tackle forecasting problems, involving univariate or multivariate data Master time series classification with residual and convolutional neural networks Get up to speed with solving time series anomaly detection problems using autoencoders and generative adversarial networks (GANs) Who this book is for If you're a machine learning enthusiast or someone who wants to learn more about building forecasting applications using deep learning, this book is for you. Basic knowledge of Python programming and machine learning is required to get the most out of this book.Table of ContentsGetting Started with Time Series Getting Started with PyTorch Univariate Time Series Forecasting Forecasting with PyTorch Lightning Global Forecasting Models Advanced Deep Learning Architectures for Time Series Forecasting Probabilistic Time Series Forecasting Deep Learning for Time Series Classification Deep Learning for Time Series Anomaly Detection

Edition details

• Publisher: Packt Publishing
• Publication date: 2024
• Language: English
• ISBN 10: 1805129236
• ISBN 13: 9781805129233
• Binding: Paperback
• Number of pages: 304