Provides a thorough coverage and comparison of a wide array of time series models and methods: Exponential Smoothing, Holt Winters, ARMA and ARIMA, deep learning models including RNNs, LSTMs, GRUs, and ensemble models composed of combinations of thes
Trade Review
"A well-structured text aimed at undergraduates pursuing a data science curriculum, or MBA students. The authors draw upon their vast combined experience in research and teaching to a variety of audiences to present the classical material on ARMA-based Box-Jenkins methodology without assuming a calculus background. Yet, their approach manages to be heuristic, while not sacrificing relevant theoretical detail that enriches understanding. The authors complement this material with chapters on multivariate models, and, refreshingly, a very enlightening discussion on neural networks. The exposition is lucid, well-organized, and copiously illustrated to reinforce comprehension of concepts. The companion R package (tswge) finds a niche in the growing list of time series toolboxes, by providing clean, straightforward functionality on such essentials as spectrum reconstruction and model factor tables to glean the structure of AR and MA polynomials."
- Alex Trindade, Texas Tech University
Table of Contents
1. Working with Data Collected Over Time, 2. Exploring Time Series Data, 3. Statistical Basics for Time Series Analysis, 4. The Frequency Domain, 5. ARMA Models, 6. ARMA Fitting and Forecasting, 7. ARIMA, Seasonal,and ARCH/GARCH Models, 8. Time Series Regression, 9. Model Assessment, 10. Multivariate Time Series, 11. Deep Neural Network Based Time Series Models
