Package Organization

The utiml package is organized in semantic groups:

• Classification Methods
• Ensemble utilities
• Evaluation measures
• Internal utilities
• Pre-process utilities
• Public utilities
• Sampling methods
• Threshold methods

Classification Methods

Each of method is organized in a single file and need have at least three methods:

• The train method that need to return a ML model
• The prediction method that must implement the S3 predict function
• The print method that must implement the S3 print function

Ensemble utilities (ensemble.R)

Ensemble utilities used to multi-label ensemble methods

##Evaluation measures (evaluate.R and measures.R) TODO

Internal utilities

Only private functions used to the package and not public

Pre-process utilities (preprocess.R)

Contains the utilities to pre-process the multi-label dataset

Public utilities

• Base methods used for transformation problem (baselearner.R)
• utitilities.R *

Sampling methods (sampling.R)

Contains the methods to split the multi-label dataset

Threshold measures (threshold.R)

Contains the methods to transform a probability prediction into a bipartition prediction