{
  "id": "2102.08993",
  "version": "v1",
  "published": "2021-02-17T19:37:35.000Z",
  "updated": "2021-02-17T19:37:35.000Z",
  "title": "Using Distance Correlation for Efficient Bayesian Optimization",
  "authors": [
    "Takuya Kanazawa"
  ],
  "comment": "10 pages",
  "categories": [
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "We propose a novel approach for Bayesian optimization, called $\\textsf{GP-DC}$, which combines Gaussian processes with distance correlation. It balances exploration and exploitation automatically, and requires no manual parameter tuning. We evaluate $\\textsf{GP-DC}$ on a number of benchmark functions and observe that it outperforms state-of-the-art methods such as $\\textsf{GP-UCB}$ and max-value entropy search, as well as the classical expected improvement heuristic. We also apply $\\textsf{GP-DC}$ to optimize sequential integral observations with a variable integration range and verify its empirical efficiency on both synthetic and real-world datasets.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2021-02-17T19:37:35.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "efficient bayesian optimization",
      "distance correlation",
      "outperforms state-of-the-art methods",
      "max-value entropy search",
      "optimize sequential integral observations"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 10,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}