utiml_stratified_split: Labelsets Stratification Create the indexes using the...

Description Usage Arguments Value References Examples

View source: R/sampling.R

Description

Labelsets Stratification Create the indexes using the Labelsets Stratification approach.

Usage

1

Arguments

mdata

A mldr dataset

r

Desired proportion of examples in each subset, r1, . . . rk

Value

A list with k disjoint indexes subsets S1, . . .Sk

References

Sechidis, K., Tsoumakas, G., & Vlahavas, I. (2011). On the stratification of multi-label data. In Proceedings of the Machine Learningand Knowledge Discovery in Databases - European Conference, ECML PKDD (pp. 145-158).

Examples

1
2
3
4
5
6
7
8
## Not run: 
# Create 3 partitions for train, validation and test
indexes <- utiml_stratified_split(emotions, c(0.6,0.1,0.3))

# Create a stratified 10-fold
indexes <- utiml_stratified_split(emotions, rep(0.1,10))

## End(Not run)

utiml documentation built on Aug. 1, 2017, 1:01 a.m.