{
  "annotation_id": "de76896c-dbf3-4620-a0e2-5958daee1d85",
  "date_created": "2026-03-02T18:01:21.510000Z",
  "date_modified": "2026-03-02T18:01:21.510000Z",
  "file_hash": "28bb34156f50af340ee47527a863979d7be23f38009182ae7baba8ae02eab9cd",
  "private": false,
  "record": {
    "abstract": "A method to approximate continuous multi-dimensional probability density\nfunctions (PDFs) using their projections and correlations is described. The\nmethod is particularly useful for event classification when estimates of\nsystematic uncertainties are required and for the application of an unbinned\nmaximum likelihood analysis when an analytic model is not available. A simple\ngoodness of fit test of the approximation can be used, and simulated event\nsamples that follow the approximate PDFs can be efficiently generated. The\nsource code for a FORTRAN-77 implementation of this method is available.",
    "arxiv_id": "physics/9805018",
    "authors": [
      "Dean Karlen"
    ],
    "categories": [
      "physics.data-an"
    ],
    "doi": "10.1063/1.168691",
    "journal_ref": "Computers in Physics, 12:4 (1998) 380.",
    "title": "Using projections and correlations to approximate probability   distributions",
    "url": "https://arxiv.org/abs/physics/9805018"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "ff99fdcd-78c8-48ea-9c77-33cd22ab5396",
    "id": "arXiv Dataset IDs",
    "type": "Model",
    "variant": "snapshot-2026-03-01",
    "version": "0.1.0"
  },
  "user_id": 1000002
}