*Time Series* is a term used to denote data indexed in time order. For example, daily metro user counts, annual revenue, etc. would all be time series data.

In Time Series analysis, the data is usually split into 3 components; the Trend, Seasonality, and the Remainder, sometimes called noise.


*An example of Time Series Decomposition, courtesy of Daitan at medium.com*

Trend represents the overall change in the data, whether it's increasing, or decreasing. It can also be of another form.

Seasonality is the frequency in which the data is said to "recur". For example, a pizza parlor selling more pizzas on weekends can be represented by a weekly seasonality.

It is also important to note that multiple seasonality compoments can be present, and there are methods to represent that, such as through the Fourier Series.

The remainder is data which can't be explained through trend and seasonality, and holds the uncertainty and random fluctuations present in the data.