{
  "id": "1511.01304",
  "version": "v1",
  "published": "2015-11-04T12:34:10.000Z",
  "updated": "2015-11-04T12:34:10.000Z",
  "title": "Dictionary descent in optimization",
  "authors": [
    "Vladimir Temlyakov"
  ],
  "comment": "arXiv admin note: text overlap with arXiv:1206.0392",
  "categories": [
    "stat.ML",
    "math.NA"
  ],
  "abstract": "The problem of convex optimization is studied. Usually in convex optimization the minimization is over a d-dimensional domain. Very often the convergence rate of an optimization algorithm depends on the dimension d. The algorithms studied in this paper utilize dictionaries instead of a canonical basis used in the coordinate descent algorithms. We show how this approach allows us to reduce dimensionality of the problem. Also, we investigate which properties of a dictionary are beneficial for the convergence rate of typical greedy-type algorithms.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2015-11-04T12:34:10.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "dictionary descent",
      "convex optimization",
      "convergence rate",
      "coordinate descent algorithms",
      "reduce dimensionality"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}