{
  "id": "2405.16594",
  "version": "v1",
  "published": "2024-05-26T15:07:16.000Z",
  "updated": "2024-05-26T15:07:16.000Z",
  "title": "Training-Conditional Coverage Bounds under Covariate Shift",
  "authors": [
    "Mehrdad Pournaderi",
    "Yu Xiang"
  ],
  "comment": "arXiv admin note: text overlap with arXiv:2404.13731",
  "categories": [
    "stat.ML",
    "cs.LG"
  ],
  "abstract": "Training-conditional coverage guarantees in conformal prediction concern the concentration of the error distribution, conditional on the training data, below some nominal level. The conformal prediction methodology has recently been generalized to the covariate shift setting, namely, the covariate distribution changes between the training and test data. In this paper, we study the training-conditional coverage properties of a range of conformal prediction methods under covariate shift via a weighted version of the Dvoretzky-Kiefer-Wolfowitz (DKW) inequality tailored for distribution change. The result for the split conformal method is almost assumption-free, while the results for the full conformal and jackknife+ methods rely on strong assumptions including the uniform stability of the training algorithm.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2024-05-26T15:07:16.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "covariate shift",
      "training-conditional coverage bounds",
      "conformal prediction methodology",
      "training-conditional coverage guarantees",
      "covariate distribution changes"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}