diff.resamples: Inferential Assessments About Model Performance

Description

Methods for making inferences about differences between models

Usage

## S3 method for class 'resamples':
diff(x, models = x$models, metric = x$metrics, 
     test = t.test, 
     confLevel = 0.95, adjustment = "bonferroni",
     ...)
## S3 method for class 'diff.resamples':
summary(object, digits = max(3, getOption("digits") - 3), ...)

Arguments

an object generated by resamples

models

a character string for which models to compare

metric

a character string for which metrics to compare

test

a function to compute differences. The output of this function should have scalar outputs called estimate and p.value

object

a object generated by diff.resamples

adjustment

any p-value adjustment method to pass to p.adjust.

confLevel

confidence level to use for dotplot.diff.resamples. See Details below.

digits

the number of significant differences to display when printing

...

further arguments to pass to test

Value

An object of class "diff.resamples" with elements:
callthe call
difsa list for each metric being compared. Each list contains a matrix with differences in columns and resamples in rows
statisticsa list of results generated by test
adjustmentthe p-value adjustment used
modelsa character string for which models were compared.
metricsa character string of performance metrics that were used
or...
An object of class "summary.diff.resamples" with elements:
callthe call
tablea list of tables that show the differences and p-values

Details

The ideas and methods here are based on Hothorn et al (2005) and Eugster et al (2008).

For each metric, all pair-wise differences are computed and tested to assess if the difference is equal to zero.

When a Bonferroni correction is used, the confidence level is changed from confLevel to 1-((1-confLevel)/p) here p is the number of pair-wise comparisons are being made. For other correction methods, no such change is used.

References

Hothorn et al. The design and analysis of benchmark experiments. Journal of Computational and Graphical Statistics (2005) vol. 14 (3) pp. 675-699

Eugster et al. Exploratory and inferential analysis of benchmark experiments. Ludwigs-Maximilians-Universitat Munchen, Department of Statistics, Tech. Rep (2008) vol. 30

Examples

Run this code

#load(url("http://caret.r-forge.r-project.org/Classification_and_Regression_Training_files/exampleModels.RData"))

resamps <- resamples(list(CART = rpartFit,
                          CondInfTree = ctreeFit,
                          MARS = earthFit))

difs <- diff(resamps)

difs

summary(difs)

Run the code above in your browser using DataLab

Data Engineering and BI courses are free this week!