The Analysis of Time Series
Description
The field of time series analysis has undergone a remarkable transformation since the publication of the seventh edition of this book. While classical statistical models such as ARIMA, state-space models, and spectral methods remain essential, the rise of artificial intelligence (AI) has introduced groundbreaking approaches to modeling, forecasting, and generating time-dependent data. This eighth edition reflects these advancements with the addition of two new chapters: Predictive AI for Time Series and Generative AI for Time Series. These chapters bridge the gap between traditional time series methods and cutting-edge AI techniques, offering readers a comprehensive and integrated perspective on the field.
Features:
· Comprehensive coverage of classical time series models, including ARIMA, state-space models, and spectral methods
· Two new chapters on predictive and generative AI, introducing cutting-edge methods like transformers, variational autoencoders, and diffusion models
· Practical examples and illustrations using R, demonstrating the application of both classical and AI-based approaches to real-world time series data
· Emphasis on the integration of classical statistical rigor with the flexibility and scalability of AI methods
· Clear explanations and intuitive insights, making advanced concepts accessible to a broad audience
· Updated content reflecting the latest developments in time series analysis, with a focus on modern, high-dimensional, and nonlinear data challenges
This eighth edition is designed for students, researchers, and practitioners in statistics, as well as in finance, economics, climate science, health, and engineering. It serves as both a foundational text for those new to time series analysis and a valuable resource for experienced analysts seeking to engage with the rapidly evolving landscape of predictive and generative AI. With its balance of theory, practical implementation, and real-world examples, the book is ideal for use in academic courses, professional training, and self-study.
More details
Other editions
Additional editions
Previous edition
Persons
Haipeng Xing is a Professor in Applied Mathematics and Statistics at the State University of New York, Stony Brook, USA, the author of three books and numerous research papers. His research interests include quantitative finance and risk management, econometrics, applied stochastic control, and sequential statistical methodology.
Chris Chatfield is a retired Reader in Statistics at the University of Bath, UK, the author of five books and numerous research papers, and an elected senior fellow of the International Institute of Forecasters.
Content
1. Introduction. 2. Basic Descriptive Techniques. 3. Some Linear Time Series Models. 4. Fitting Time Series Models in the Time Domain. 5. Forecasting. 6. Stationary Processes in the Frequency Domain. 7. Spectral Analysis. 8. Bivariate Processes. 9. Linear Systems. 10. State-Space Models and the Kalman Filter. 11. Non-Linear Models. 12. Volatility Models. 13. Multivariate Time Series Modelling. 14. Predictive AI for Time Series. 15. Generative AI for Time Series.