# `uncertainty`

Uncertainty.

Uncertainty methods say how sure a learner is about a prediction, rather than
giving a single value on its own. In data stream learning these estimates are
updated as each new instance arrives, so they keep working when the underlying
distribution drifts.

This module currently provides prediction intervals for regression. A
prediction interval gives a range of plausible values around a point
prediction.

## Classes

| [`MVE`](capymoa.uncertainty.MVE.md#capymoa.uncertainty.MVE)     |    |
|----------------------------------------------------------------------------------|----|
| [`AdaPI`](capymoa.uncertainty.AdaPI.md#capymoa.uncertainty.AdaPI) |    |