Main Content

training

Training indices for time series cross-validation

    Description

    example

    idx = training(c) returns the training indices idx for a tspartition object c of type 'holdout'. That is, the logical vector idx specifies the observations in the training set.

    example

    idx = training(c,i) returns the training indices for window i of a tspartition object c of type 'expanding-window' or 'sliding-window'. That is, the logical vector idx specifies the observations in training set i.

    • If c.Type is 'expanding-window', then the training set size expands with each window while the test set size remains fixed.

    • If c.Type is 'sliding-window', then both the training set size and the test set size are fixed.

    Examples

    collapse all

    Identify the observations in the training set of a tspartition object for holdout validation.

    Use 30% of 20 time-dependent observations to create a test set. The corresponding training set contains the remaining observations.

    c = tspartition(20,"Holdout",0.30);

    Find the training set indices. A value of 1 (true) indicates that the corresponding observation is in the training set. A value of 0 (false) indicates that the corresponding observation is in the test set.

    trainingIndices = training(c);

    Visualize the observations in the training set by using a heat map.

    h = heatmap(double(trainingIndices),ColorbarVisible="off");
    h.XDisplayLabels = "";
    ylabel("Observation")
    title("Training Set Observations")

    Figure contains an object of type heatmap. The chart of type heatmap has title Training Set Observations.

    The observations in dark blue (with a value of 1) are in the training set, and the observations in light blue (with a value of 0) are in the test set. When you use holdout validation for time series data, the latest observations (in this case, observations 15 through 20) are in the test set.

    Identify the observations in the training sets and test sets of a tspartition object for expanding window cross-validation.

    Use 20 time-dependent observations to create three training sets and three test sets. Specify a gap of two observations between each training set and its corresponding test set.

    c = tspartition(20,"ExpandingWindow",3, ...
        GapSize=2);

    Find the training set indices for the three windows. A value of 1 (true) indicates that the corresponding observation is in the training set for that window.

    trainWindow1 = training(c,1);
    trainWindow2 = training(c,2);
    trainWindow3 = training(c,3);

    Find the test set indices for the three windows. A value of 1 (true) indicates that the corresponding observation is in the test set for that window.

    testWindow1 = test(c,1);
    testWindow2 = test(c,2);
    testWindow3 = test(c,3);

    Combine the training and test set indices into one matrix where a value of 1 indicates a training observation and a value of 2 indicates a test observation.

    data = [trainWindow1 + 2*testWindow1, ...
        trainWindow2 + 2*testWindow2, ...
        trainWindow3 + 2*testWindow3];

    Visualize the different sets by using a heat map.

    colormap = lines(3);
    heatmap(double(data),ColorbarVisible="off", ...
        Colormap=colormap);
    xlabel("Window")
    ylabel("Observation")
    title("Expanding Window Cross-Validation Scheme")

    Figure contains an object of type heatmap. The chart of type heatmap has title Expanding Window Cross-Validation Scheme.

    For each window, the observations in red (with a value of 1) are in the training set, the observations in yellow (with a value of 2) are in the test set, and the observations in blue (with a value of 0) are ignored. For example, observation 11 is a test observation in window one, a gap observation in window two, and a training observation in window three.

    Input Arguments

    collapse all

    Time series validation partition, specified as a tspartition object. The validation partition type (Type) is 'expanding-window', 'holdout', or 'sliding-window'.

    Training set or window index, specified as a positive integer scalar. When you specify i, the training function finds the observations in training set i.

    Data Types: single | double

    Output Arguments

    collapse all

    Indices for training set observations, returned as a logical vector. A value of 1 (true) indicates that the corresponding observation is in the training set. A value of 0 (false) indicates that the corresponding observation is in a different set, such as the test set.

    Version History

    Introduced in R2022b