Summary
Preprocessing is a crucial step in machine learning that is often neglected. Many books don't cover preprocessing as a topic or skip preprocessing entirely. However, it is often in preprocessing that relatively easy wins can be achieved. The quality of the data determines the outcome.
Preprocessing includes curating and screening the data. The expected output of the preprocessing is a dataset on which it is easier to conduct machine learning. This can mean that it is more reliable and less noisy than the original dataset.
We've talked about feature transforms and feature engineering approaches to time-series data, and we've talked about automated approaches as well.
In the next chapters, we'll explore how we can use these extracted features in a machine learning model. We'll discuss combinations of features and modeling algorithms in the next chapter, Chapter 4, Introduction to Machine Learning for Time-Series. In Chapter 5, Time-Series...
