You're reading from Modern Time Series Forecasting with Python Industry-ready machine learning and deep learning time series analysis with PyTorch and pandas

Product type Paperback

Published in Oct 2024

Publisher Packt

ISBN-13 9781835883181

Length 658 pages

Edition 2nd Edition

Languages

Python

Tools

PyTorch

Concepts

Data Science

Authors (2):

Jeffrey Tackes

Manu Joseph

View More author details

Table of Contents (26) Chapters

Preface

1. Part-1: Getting Familiar with Time Series FREE CHAPTER

2. Introducing Time Series

3. Acquiring and Processing Time Series Data

4. Analyzing and Visualizing Time Series Data

5. Setting a Strong Baseline Forecast

6. Part-2: Machine Learning for Time Series

7. Time Series Forecasting as Regression

8. Feature Engineering for Time Series Forecasting

9. Target Transformations for Time Series Forecasting

10. Forecasting Time Series with Machine Learning Models

11. Ensembling and Stacking

12. Global Forecasting Models

13. Part-3: Deep Learning for Time Series

14. Introduction to Deep Learning

15. Building Blocks of Deep Learning for Time Series

16. Common Modeling Patterns for Time Series

17. Attention and Transformers for Time Series

18. Strategies for Global Deep Learning Forecasting Models

19. Specialized Deep Learning Architectures for Forecasting

20. Probabilistic Forecasting and More

21. Part-4: Mechanics of Forecasting

22. Multi-Step Forecasting

23. Evaluating Forecast Errors—A Survey of Forecast Metrics

24. Evaluating Forecasts—Validation Strategies

25. Index

Time Series Dense Encoder (TiDE)

We saw earlier in the chapter that a linear family of models outperformed quite a lot of Transformer models. In 2023, Das et al. from Google proposed a model that extends that idea into non-linearity. They argued that the linear models will fall short where there are inherent non-linearities in the dependence between the future and the past. The inclusion of covariates compounds this problem.

Reference check:

The research paper by Das et al. on TiDE is cited in the References section as 20.

Therefore, they introduced a simple and efficient Multi-Layer Perceptron (MLP) based architecture for long-term time series forecasting. The model essentially encodes the past of the time series, along with the covariates using dense MLPs, and then decodes this latent representation into a forecast. The model assumes channel independence (similar to PatchTST) and considers different related time series in a multivariate problem as separate...

The rest of the chapter is locked

You're reading from Modern Time Series Forecasting with Python Industry-ready machine learning and deep learning time series analysis with PyTorch and pandas

Table of Contents (26) Chapters

Time Series Dense Encoder (TiDE)

Unlock this book and the full library FREE for 7 days

Authors (2)

Personalised recommendations for you