{
  "annotation_id": "a3f86605-c89f-4de9-a8c1-1154a4250453",
  "date_created": "2026-03-02T18:00:53.502000Z",
  "date_modified": "2026-03-02T18:00:53.502000Z",
  "file_hash": "71f7ae3ebebe6f41aa0eb25a9d302c33e8c6a199ce6dbee9a5608523332f805b",
  "private": false,
  "record": {
    "abstract": "A recently proposed mutual information based algorithm for decomposing data\ninto least dependent components (MILCA) is applied to spectral analysis, namely\nto blind recovery of concentrations and pure spectra from their linear\nmixtures. The algorithm is based on precise estimates of mutual information\nbetween measured spectra, which allows to assess and make use of actual\nstatistical dependencies between them. We show that linear filtering performed\nby taking second derivatives effectively reduces the dependencies caused by\noverlapping spectral bands and, thereby, assists resolving pure spectra. In\ncombination with second derivative preprocessing and alternating least squares\npostprocessing, MILCA shows decomposition performance comparable with or\nsuperior to specialized chemometrics algorithms. The results are illustrated on\na number of simulated and experimental (infrared and Raman) mixture problems,\nincluding spectroscopy of complex biological materials.\n  MILCA is available online at http://www.fz-juelich.de/nic/cs/software",
    "arxiv_id": "physics/0412029",
    "authors": [
      "Sergey A. Astakhov",
      "Harald St\u00f6gbauer",
      "Alexander Kraskov",
      "Peter Grassberger"
    ],
    "categories": [
      "physics.data-an",
      "cs.IT",
      "math.IT",
      "physics.chem-ph"
    ],
    "title": "Spectral Mixture Decomposition by Least Dependent Component Analysis",
    "url": "https://arxiv.org/abs/physics/0412029"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "caf84ece-5439-45bc-9591-01b3bf234201",
    "id": "arXiv Dataset IDs",
    "type": "Model",
    "variant": "snapshot-2026-03-01",
    "version": "0.1.0"
  },
  "user_id": 1000002
}