Machine Learning for Time-Series with Python: Forecast, predict, and detect anomalies with state-of-the-art machine learning methods

by Ben Auffarth

Get better insights from time-series data and become proficient in model performance analysis

Key FeaturesExplore popular and modern machine learning methods including the latest online and deep learning algorithmsLearn to increase the accuracy of your predictions by matching the right model with the right problemMaster time series via real-world case studies on operations management, digital marketing, finance, and healthcareBook DescriptionThe Python time-series ecosystem is huge and often quite hard to get a good grasp on, especially for time-series since there are so many new libraries and new models. This book aims to deepen your understanding of time series by providing a comprehensive overview of popular Python time-series packages and help you build better predictive systems.

Machine Learning for Time-Series with Python starts by re-introducing the basics of time series and then builds your understanding of traditional autoregressive models as well as modern non-parametric models. By observing practical examples and the theory behind them, you will become confident with loading time-series datasets from any source, deep learning models like recurrent neural networks and causal convolutional network models, and gradient boosting with feature engineering.

This book will also guide you in matching the right model to the right problem by explaining the theory behind several useful models. You’ll also have a look at real-world case studies covering weather, traffic, biking, and stock market data.

By the end of this book, you should feel at home with effectively analyzing and applying machine learning methods to time-series.

What you will learnUnderstand the main classes of time series and learn how to detect outliers and patternsChoose the right method to solve time-series problemsCharacterize seasonal and correlation patterns through autocorrelation and statistical techniquesGet to grips with time-series data visualizationUnderstand classical time-series models like ARMA and ARIMAImplement deep learning models, like Gaussian processes, transformers, and state-of-the-art machine learning modelsBecome familiar with many libraries like Prophet, XGboost, and TensorFlowWho this book is forThis book is ideal for data analysts, data scientists, and Python developers who want instantly useful and practical recipes to implement today, and a comprehensive reference book for tomorrow. Basic knowledge of the Python Programming language is a must, while familiarity with statistics will help you get the most out of this book.

Table of ContentsIntroduction to Time-Series with PythonTime-Series Analysis with PythonPreprocessing Time-SeriesIntroduction to Machine Learning for Time SeriesForecasting with Moving Averages and Autoregressive ModelsUnsupervised Methods for Time-SeriesMachine Learning Models for Time-SeriesOnline Learning for Time-SeriesProbabilistic Models for Time-SeriesDeep Learning for Time-SeriesReinforcement Learning for Time-SeriesMultivariate Forecasting

370 pages, Kindle Edition

Published October 29, 2021

Reviews

[Ondrej · Rating 4 out of 5]

This book does a good job of capturing the state-of-the-art techniques for time series in every major area of Machine Learning into 300 pages.

Every chapter begins with a short introduction of the theory about the topic and methods. As every chapter covers a relatively broad field, do not expect exhaustive in-depth explanations for every algorithm but rather a quick overview of all the essentials you need to know to start. I genuinely enjoyed this approach as it gives enough information about the methods (which readers can explore further in original articles) while not being too long for someone already skilled in the topic.

Afterward, there is an outline of the most used python packages, which you can use to try out the theory from the chapter. It consists mainly of short code snippets, and I would compare it to assembled "Getting Started" sections from package documentation. I think it accomplishes the intended purpose of giving you a short example of how to use a particular package, but it does not go much further.

Even though prior knowledge about the topic is required, I would recommend this book to anyone looking for a starting point in machine learning for time series. One thing I was missing were more detailed comparisons of mentioned methods, their advantages, when and why to use them, but maybe that would require a book on its own.

[Matthew Mechtly · Rating 2 out of 5]

While much of the non-coding material will remain relevant for many years, many of the coding portions of the book were non-functional less than a year after publication when I initially tried to get them to work. The author says he keeps his GitHub page up-to-date with the most recent code changes, but this is simply not true.

In the future, I'd recommend the author include environment requirement files to ensure environments are replicable for users several months after publication.