Practical Time Series Analysis: Prediction with Statistics and Machine Learning

by Aileen Nielsen

Time series data analysis is increasingly important due to the massive production of such data through the internet of things, the digitalization of healthcare, and the rise of smart cities. As continuous monitoring and data collection become more common, the need for competent time series analysis with both statistical and machine learning techniques will increase. Covering innovations in time series data analysis and use cases from the real world, this practical guide will help you solve the most common data engineering and analysis challenges in time series, using both traditional statistical and modern machine learning techniques. Author Aileen Nielsen offers an accessible, well-rounded introduction to time series in both R and Python that will have data scientists, software engineers, and researchers up and running quickly. You’ll get the guidance you need to

Genres: Programming, Nonfiction, Technical, Artificial Intelligence, Technology, Computer Science

497 pages, Paperback

Published November 19, 2019

Reviews

[Bojan Tunguz · Rating 3 out of 5]

This is a very "big picture" book on modern time series analysis as it is practiced in Data Science and related domains. It gives a general overview of the main techniques, problems, and issues that arise in this field. However, from my standpoint, it is as far removed from the "practical" introduction as a book of this kind can be. There are very few worked out examples, and most of the techniques in the book are not as up-to-date as one would have liked. This is unfortunate, as the time-series analysis and predictive modeling are very hot topics, and there are innumerable practical applications in almost any area of modern data science. The book provides a decent general overviews, but in order to learn anything truly applied and practical, I would recommend that one looks at many good online tutorials. In particular, I'd strongly recommend taking a look at the time-series Kaggle competitions.

[Matt Aadland · Rating 4 out of 5]

I think this is a good introductory book to learn the basics of time series analysis and it would be a good companion to "Forecasting: Principles and Practice" by Rob J Hyndman and George Athanasopoulos

If you were to only read one book, I'd pick Forecasting: Principles and Practice as it goes into much more foundational detail.

[Hồ Vinh · Rating 4 out of 5]

This is my first book to further broaden my knowledge on the time series topic methodically. I forced myself through it and concluded this was a mere collection of discrete ideas vaguely connected altogether, especially its strangely reiterated lookahead concept - another name for the data leak issue in training. Then, I partook in a time series forecast use case, and the book as a reference note served me well, and I appreciated it a bit more.

To sum up, the content is structured to follow an end-to-end time series development journey. It discusses the best practices of EDA, feature engineering, modelling, and data storing with valuable tips: lookahead issue, plotting techniques, temporal characteristics in analysis, etc... Essentially, all the relevant concepts for beginners.

However, the mathematical explanation is not the author's strong suit, hence it's more complex than it is and looking up online alternatives is better in general. But if you can skim through that and absorb the practice guideline, plus actual hands-on, it's a good companion book.

Here are what I learned:
- Visualizing data with 2 temporal axes to discover seasonality.
- Adjust your performance expectation by comparing the time series model with an acceptable baseline (e.g. yesterday's value is today's forecast).
- Lookahead is often disguised in data processing; for example, a smoothing function may include future data in its computation.
- Formulate to forecast the movement instead of the amount of the output.
- Backtesting with rolling window.
- Handpick the process formulation by studying the ACF and PACF behaviour.

[Chris King · Rating 5 out of 5]

Read this if you are looking to learn about time series models.

[Xingda Wang · Rating 4 out of 5]

Read until page 190, I find it somehow not clear especially on how model is setup and criterion analysis.

[Abhijeet Upadhyay · Rating 1 out of 5]

Shallow explanations and multiple errors. Not useful.