{
  "id": "1709.01632",
  "version": "v1",
  "published": "2017-09-05T23:52:14.000Z",
  "updated": "2017-09-05T23:52:14.000Z",
  "title": "Mean-field theory of Bayesian clustering",
  "authors": [
    "Alexander Mozeika",
    "Anthony CC Coolen"
  ],
  "categories": [
    "cond-mat.dis-nn",
    "physics.data-an"
  ],
  "abstract": "We map model-based Bayesian clustering, which is based on stochastic partitioning of data, into a statistical physics problem for a gas of particles. Using mean-field theory we show that, under natural assumptions, the lowest entropy state of this hypothetical `gas' corresponds to the optimal clustering of data. The byproduct of our analysis is a simple but effective clustering algorithm, which infers both the most plausible number of clusters and the corresponding partitions.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2017-09-05T23:52:14.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "mean-field theory",
      "lowest entropy state",
      "natural assumptions",
      "statistical physics problem",
      "map model-based bayesian clustering"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}