An objective evaluation criterion for clustering

Arindam Banerjee, John LangfordKDD 2004 · OpenReview

Abstract

We propose and test an objective criterion for evaluation of clustering performance: How well does a clustering algorithm run on unlabeled data aid a classification algorithm? The accuracy is quantified using the PAC-MDL bound [3] in a semisupervised setting. Clustering algorithms which naturally separate the data according to (hidden) labels with a small number of clusters perform well. A simple extension of the argument leads to an objective model selection method. Experimental results on text analysis datasets demonstrate that this approach empirically results in very competitive bounds on test set performance on natural datasets.
BibTeX
@inproceedings{banerjee2004objective,
  author = {Arindam Banerjee and John Langford},
  title = {{An objective evaluation criterion for clustering}},
  booktitle = {ACM SIGKDD Conference on Knowledge Discovery and Data Mining},
  year = {2004},
  url = {https://openreview.net/forum?id=SyNBC4-uZH},
}

Loading related content...

Figure