What's next for time-series?
We've looked at many aspects of time-series in this book. If you've made it this far, you should have learned how to analyze time-series, and how to apply traditional time-series forecasts. This is often the main focus of other books on the market; however, we went far beyond.
We looked at preprocessing and transformations for time-series as relevant to machine learning. We looked at many examples of applying machine learning both in an unsupervised and supervised context for forecasting and other predictions, anomaly detection, and drift and change point detection. We delved into techniques such as online learning, reinforcement learning, probabilistic models, and deep learning.
In each chapter, we've been looking at the most important libraries, sometimes even the cutting edge, and, finally, prevalent industrial applications. We've looked at state-of-the-art models such as HIVE-COTE, preprocessing methods such as ROCKET...
