MAE: The Mean Absolute Error

In SurvMetrics: Predictive Evaluation Metrics in Survival Analysis

The Mean Absolute Error

Description

A somewhat naive criterion that is sometimes used consists of simply omitting all censored cases from the data set. For survival analysis problems, the mean absolute error (MAE) can be defined as an average of the differences between the predicted time values and the actual observation time values. Only the samples for which the event occurs are being considered in this metric.

Usage

MAE(object, pre_time)

Arguments

object	object of class Surv created by Surv function.
pre_time	a vector of predicted values of survival time of each observation.

Details

Condition: MAE can only be used for the evaluation of survival models which can provide the event time as the predicted target value.

Value

the value of the Mean Absolute Error between the survival time and the predicted result.

Author(s)

Hanpu Zhou [e-mail]

References

Matsuo, K. , Purushotham, S. , Jiang, B. , Mandelbaum, R. S. , Takiuchi, T. , & Liu, Y. , et al. (2018). Survival outcome prediction in cervical cancer: cox models vs deep-learning model. American Journal of Obstetrics & Gynecology. Coyle, E. J. , & Lin, J. H. . (1988). Stack filters and the mean absolute error criterion. IEEE Trans Acoustics Speech Signal Processing, 36(8), 1244-1254.

Examples

library(survival)
time <- rexp(50)
status <- sample(c(0, 1), 50, replace = TRUE)
pre_time <- rexp(50)
MAE(Surv(time, status), pre_time)

SurvMetrics documentation built on Sept. 4, 2022, 1:06 a.m.