Chapter 9. Understanding Sequence and Time Series Data
Time series are everywhere. You’ve probably seen them in things like weather forecasts, stock prices, and historic trends like Moore’s law. If you’re not familiar with Moore’s law, it predicts that the number of transistors on a microchip will roughly double every two years—and for almost 50 years, it has proven to be an accurate predictor of the future of computing power and cost (see Figure 9-1).
Figure 9-1. Moore’s law
Note
The gaps in Figure 9-1 are missing data for that period of time, but the general trend still holds.
Time series data is a set of values that are spaced over time, usually in a particular order or denoting values of a thing at a timestamped point in time. When a time series is plotted on a graph, the x-axis is usually temporal in nature. Often, there are a number of values plotted on the time axis, such as in the example shown in Figure 9-1, where the number of transistors is one plot and the predicted value from Moore’s law is the other. This is called a multivariate time series. If there’s just a single value—for example, the volume of rainfall over time—then it’s called a univariate time series.
With Moore’s law, predictions are simple because there’s a fixed and simple rule that allows us to roughly predict the future—a rule that has held for about 50 years.
But what about a time series like the ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access