Outliers
These are data points that differ significantly from other values in the dataset. Outliers may result from measurement or data entry errors, or they may reflect natural variations.
To learn more, see Insights.

Related data (drivers)
An optional dataset that provides additional drivers (business factors and values) to your forecast. Related data can help Forecaster to improve forecast accuracy. Examples include historical data and forward-looking promotions. Related data include:

A time dimension (for example: historical and forward-looking promotions).
Historical data points leading up to the forecast horizon.
Future data points to cover the length of the forecast horizon (recommended for increased forecast accuracy).

Sparse data
A dataset where many of the values are zero. For example, if your dataset represents demand for a product over time, and there are periods where demand is zero. You can use the sparsity information you get from Forecaster Insights to check your data.

Before you include a dataset in a data collection, you can choose to distinguish between true zeros and zeros that represent missing data. This distinction helps ensure that your forecast results aren't skewed. For more information, see Exclude values.

Sparsity
This term refers to how much of the data is populated with meaningful values. High sparsity indicates that a large portion of the data is missing or consists of zeros.

For more details about Sparsity, see Insights.

Seasonality
This type of data offers insights into cyclical, repeated patterns in historical data that occur at regular intervals. Forecaster detects seasonality in the target historical data. Seasonality depends on the data collection periodicity (daily, weekly, monthly, yearly).

For more information, see Insights.

Summary statistics
Summary or descriptive statistics recap a dataset's main characteristics. They provide a summary of the overall distribution and structure of the data.

For details about summary statistics, see Insights.

Time series
A sequence of successive data points (observations) in time, that occur at regular time intervals.

Time series forecasting
Prediction of a time series' future values based on historical and other data types.

Trend
This refers to the overall direction of historical data, calculated over a quarterly or yearly basis. It shows any detected trends in the target historical data. Example: a three-month product sale could result in a quarterly sales trend.

For more details, see Insights.

Variability type
This refers to a combination of the variation in timing, and the variation of quantity in historical data.

Smooth has low variability in timing and quantity.
Lumpy has high variability in timing and quantity.
Intermittent has high variability in timing and low variability in quantity.
Erratic has low variability in timing and high variability in quantity.
Unknown has insufficient data to calculate variability type.

For more information, see Insights.

title