This function calculates the interval score for a given set of
observations, lower and upper bounds, and alpha parameter.
Usage
int_score(y, l, u, alpha)
Arguments
- y
A numeric vector of observations.
- l
A numeric vector of lower bounds for the prediction intervals.
- u
A numeric vector of upper bounds for the prediction intervals.
- alpha
A numeric value specifying the penalty parameter for interval
width.
Value
A numeric vector of interval scores.
Details
The interval score is a proper scoring rule that measures the
accuracy of interval predictions. It takes into account both the coverage
and the width of the prediction interval. A lower score indicates a better
prediction.