MLPUGS-package: MLPUGS: Multi-Label Prediction Using Gibbs Sampling (and...

In bearloga/MLPUGS: Multi-Label Prediction Using Gibbs Sampling (and Classifier Chains)

Description

An implementation of classifier chains for binary and probabilistic multi-label prediction. The classification pipeline consists of:

1. Training an ensemble of classifier chains. Each chain is a binary classifier (built-in, supplied from an external package or user-coded).

2. Making predictions using a Gibbs sampler since each unobserved label is conditioned on the others.

3. (Optional) Evaluating the ECC.

4. Gathering predictions (aggregating across iterations & models).

To learn more about MLPUGS, start with the vignettes: browseVignettes(package = "MLPUGS")

bearloga/MLPUGS documentation built on March 7, 2020, 11:02 a.m.