Wow! eBook
Modern Time Series Forecasting with Python: Explore industry-ready time series forecasting using modern machine learning and deep learning
Author(s): Manu Joseph (Author)
- Publisher: Packt Publishing
- Publication Date: 24 Nov. 2022
- Language: English
- Print length: 552 pages
- ISBN-10: 1803246804
- ISBN-13: 9781803246802
Book Description
Build real-world time series forecasting systems which scale to millions of time series by applying modern machine learning and deep learning concepts
Key Features
- Explore industry-tested machine learning techniques used to forecast millions of time series
- Get started with the revolutionary paradigm of global forecasting models
- Get to grips with new concepts by applying them to real-world datasets of energy forecasting
Book Description
We live in a serendipitous era where the explosion in the quantum of data collected and a renewed interest in data-driven techniques such as machine learning (ML), has changed the landscape of analytics, and with it, time series forecasting. This book, filled with industry-tested tips and tricks, takes you beyond commonly used classical statistical methods such as ARIMA and introduces to you the latest techniques from the world of ML.
This is a comprehensive guide to analyzing, visualizing, and creating state-of-the-art forecasting systems, complete with common topics such as ML and deep learning (DL) as well as rarely touched-upon topics such as global forecasting models, cross-validation strategies, and forecast metrics. You’ll begin by exploring the basics of data handling, data visualization, and classical statistical methods before moving on to ML and DL models for time series forecasting. This book takes you on a hands-on journey in which you’ll develop state-of-the-art ML (linear regression to gradient-boosted trees) and DL (feed-forward neural networks, LSTMs, and transformers) models on a real-world dataset along with exploring practical topics such as interpretability.
By the end of this book, you’ll be able to build world-class time series forecasting systems and tackle problems in the real world.
What you will learn
- Find out how to manipulate and visualize time series data like a pro
- Set strong baselines with popular models such as ARIMA
- Discover how time series forecasting can be cast as regression
- Engineer features for machine learning models for forecasting
- Explore the exciting world of ensembling and stacking models
- Get to grips with the global forecasting paradigm
- Understand and apply state-of-the-art DL models such as N-BEATS and Autoformer
- Explore multi-step forecasting and cross-validation strategies
Who this book is for
The book is for data scientists, data analysts, machine learning engineers, and Python developers who want to build industry-ready time series models. Since the book explains most concepts from the ground up, basic proficiency in Python is all you need. Prior understanding of machine learning or forecasting will help speed up your learning. For experienced machine learning and forecasting practitioners, this book has a lot to offer in terms of advanced techniques and traversing the latest research frontiers in time series forecasting.
Table of Contents
- Introducing Time Series
- Acquiring and Processing Time Series Data
- Analyzing and Visualizing Time Series Data
- Setting a Strong Baseline Forecast
- Time Series Forecasting as Regression
- Feature Engineering for Time Series Forecasting
- Target Transformations for Time Series Forecasting
- Forecasting Time Series with Machine Learning Models
- Ensembling and Stacking
- Global Forecasting Models
- Introduction to Deep Learning
- Building Blocks of Deep Learning for Time Series
- Common Modeling Patterns for Time Series
- Attention and Transformers for Time Series
- Strategies for Global Deep Learning Forecasting Models
(N.B. Please use the Look Inside option to see further chapters)
Editorial Reviews
Review
“Manu Joseph has delivered a book full of actionable ideas and code snippets. This book will help any data scientist to get up to speed with forecasting models in no time.”
—
Nicolas Vandeput, author of Dasta Science for Supply Chain Forecasting
“Modern Time Series Forecasting with Python is a comprehensive guide to forecasting time series data using both statistical and machine learning techniques. The book strikes an effective balance between theory and practical application, making advanced forecasting techniques accessible to readers.
The book starts with a strong theoretical foundation, introducing key time series concepts like stationarity, seasonality, autocorrelation, and visualizations. Mathematical notation is clean and consistent without overwhelming the beginner. The inclusion of synthetic data generation using tools like TimeSynth gives useful hands-on practice.
A highlight is the emphasis on real-world application using a smart meter dataset. The thorough data processing chapter equips readers to handle messy data with techniques like interpolation to fill gaps. Classifiers like isolation forest and S-ESD for automatic outlier detection are cutting-edge. The book then ably crosses the bridge from classical to modern techniques.
The machine learning section starts gently, easing the reader into fundamental concepts before demonstrating linear regression, random forests, and gradient boosting for time series problems with exceptional clarity. The standardized code framework developed allows convenient benchmarking.
Thereafter, the book thoroughly grounds the reader in deep learning fundamentals like loss functions, activation, and optimization before launching into specialized architectures. The coverage of encoder-decoder patterns, attention mechanisms, and transformers is comprehensive yet accessible. The author strikes an admirable balance between theory and practical application throughout.
The book concludes strongly with the rarely discussed topics of multi-step forecasting, appropriate error metrics, and validation strategies. The mechanism for mathematical optimization of metrics stands out. Discussions are evidence-based with empirical studies and thoughtful guidelines provided.
MDTSwP is an exceptional resource for anyone looking to advance their data science skills. The mathematical depth and programming detail may challenge beginners, but perseverance will be rewarded. For intermediate practitioners, this book elevates forecasting technique to an advanced level through both classical and cutting-edge approaches delivered with remarkable clarity.”
—
Lan Laucirica
CEO | Prime Deviation
“This book is an excellent companion for data practitioners working with time series data. The author provides a comprehensive understanding into time series techniques, starting from the fundamentals to more recent and advanced techniques using transformer-based techniques. The writing and explanations are simple, making the book accessible for a large audience.”
—
Anirban Sengupta| Sr. Manager of Applied Science| Amazon
About the Author
Manu Joseph is a self-made data scientist with more than a decade of experience working with many Fortune 500 companies enabling digital and AI transformations, specifically in machine learning-based demand forecasting. He is considered an expert, thought leader, and strong voice in the world of time series forecasting. Currently, Manu leads applied research at Thoucentric, where he advances research by bringing cutting-edge AI technologies to the industry. He is also an active open-source contributor and developed an open-source library―PyTorch Tabular―which makes deep learning for tabular data easy and accessible. Originally from Thiruvananthapuram, India, Manu currently resides in Bengaluru, India, with his wife and son
