{
  "id": "2007.03511",
  "version": "v1",
  "published": "2020-07-06T17:21:24.000Z",
  "updated": "2020-07-06T17:21:24.000Z",
  "title": "Estimating Generalization under Distribution Shifts via Domain-Invariant Representations",
  "authors": [
    "Ching-Yao Chuang",
    "Antonio Torralba",
    "Stefanie Jegelka"
  ],
  "comment": "arXiv admin note: text overlap with arXiv:1910.05804",
  "journal": "International Conference on Machine Learning, 2020",
  "categories": [
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "When machine learning models are deployed on a test distribution different from the training distribution, they can perform poorly, but overestimate their performance. In this work, we aim to better estimate a model's performance under distribution shift, without supervision. To do so, we use a set of domain-invariant predictors as a proxy for the unknown, true target labels. Since the error of the resulting risk estimate depends on the target risk of the proxy model, we study generalization of domain-invariant representations and show that the complexity of the latent representation has a significant influence on the target risk. Empirically, our approach (1) enables self-tuning of domain adaptation models, and (2) accurately estimates the target error of given models under distribution shift. Other applications include model selection, deciding early stopping and error detection.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2020-07-06T17:21:24.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "distribution shift",
      "domain-invariant representations",
      "estimating generalization",
      "target risk",
      "domain adaptation models"
    ],
    "tags": [
      "journal article"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}