Contents

TimeSeriesForecasterAnnotatedWindows

A sequence of forecasting windows on a time series shaped array.

Declaration

struct TimeSeriesForecasterAnnotatedWindows<Scalar> where Scalar : MLShapedArrayScalar

Overview

A time-series forecaster takes a series of samples and produces a prediction of the next samples. For example the sequence [1, 2, 3, 4] could predict [5, 6].

The shape of each feature in the sequence is [inputWindowSize, featureSize] and the shape of each annotation is [forecastWindowSize, annotationSize]. The sequence will return as many feature-annotation examples as fit in the input. For example an input sequence of size of 10 with an input sample count of 4, a prediction sample count of 2, and a stride of 1 will produce 5 annotated windows:

feature: [1, 2, 3, 4], annotation: [5, 6]
feature: [2, 3, 4, 5], annotation: [6, 7]
feature: [3, 4, 5, 6], annotation: [7, 8]
feature: [4, 5, 6, 7], annotation: [8, 9]
feature: [5, 6, 7, 8], annotation: [9, 10]

Note that 9 and 10 are never used as features because there would be no annotations for those samples.

Topics

Creating a time series forecaster annotated window

Inspecting a time series forecaster annotated window

Default Implementations

See Also

Time-based components