estimationSummary: Obtain a set of descriptive statistics of the scores of a...

In performanceEstimation: An Infra-Structure for Performance Estimation of Predictive Models

Description

This function provides a set of descriptive statistics for each evaluation metric that is estimated on a performance estimation comparison. These statistics are obtained for a particular workflow, and for one of the prediction tasks involved in the experiment.

Usage

estimationSummary(results,workflow,task)

Arguments

results	This is a ComparisonResults object (type "class?ComparisonResults" for details) that contains the results of a performance estimation experiment obtained through the performanceEstimation() function.
workflow	A string with the ID of a workflow (it can also be an integer).
task	A string with the ID of a task (it can also be an integer).

Value

The function returns a matrix with the rows representing summary statistics of the scores obtained by the model on the different iterations, and the columns representing the evaluation statistics estimated in the experiment.

Author(s)

Luis Torgo [e-mail]

References

Torgo, L. (2014) An Infra-Structure for Performance Estimation and Experimental Comparison of Predictive Models in R. arXiv:1412.0436 [cs.MS] http://arxiv.org/abs/1412.0436

See Also

getScores, performanceEstimation

Examples

## Not run: 
## Estimating MSE for 3 variants of both
## regression trees and SVMs, on  two data sets, using one repetition
## of 10-fold CV
library(e1071)
data(swiss)

## running the estimation experiment
res <- performanceEstimation(
  PredTask(Infant.Mortality ~ .,swiss),
  workflowVariants(learner="svm",
                   learner.pars=list(cost=c(1,10),gamma=c(0.01,0.5))),
  EstimationTask("mse",method=CV(nReps=2,nFolds=5))
  )

## Get the summary of the estimations of svm.v2 on swiss
estimationSummary(res,"svm.v2","swiss.Infant.Mortality")


## End(Not run)

performanceEstimation documentation built on May 2, 2019, 6:01 a.m.