What Is a Time Series and How Do We Analyze Its Patterns?

Time Series

A time series is a sequence of data points ordered chronologically, varying over time and interrelated. Observing and studying a time series to discover its development patterns and forecast its future trajectory constitutes time series analysis. Depending on the focus of study, time series can be classified in several ways.

Univariate vs. multivariate: Based on the number of variables examined, there are univariate time series and multivariate time series.

Discrete vs. continuous: According to the continuity of time, time series are divided into discrete-time series and continuous-time series.

Stationary vs. non‑stationary: Based on statistical properties, there are stationary and non‑stationary series. A series whose probability distribution does not depend on time is called a strict (or narrow) stationary series. If the first and second moments exist and for any time the mean is constant and the covariance depends only on the time lag, the series is called weakly stationary (also known as wide‑sense or covariance‑stationary).

Gaussian vs. non‑Gaussian: According to the distributional pattern, series can be Gaussian (normal) or non‑Gaussian.