{
  "abstract": "Effective unsupervised domain adaptation (UDA) requires a representation that captures transferable structure from both the labeled source domain and the unlabeled target domain. We introduce Joint Contrastive-Adversarial Representation Learning (JCAR), a two-stage framework that learns a common representation from the union of unlabeled source and target samples by combining contrastive learning with adversarial domain alignment, and then freezes the representation to fit a source-supervised predictor. We establish a non-asymptotic excess-risk bound for this procedure that couples representation and domain-discriminator estimation while separating downstream prediction, with explicit neural-network approximation errors. Under the stated assumptions and common Hölder smoothness $β\\geq1$, the bound identifies a sufficiently large unlabeled-sample regime in which the downstream rate is $\\widetilde{\\mathcal{O}}(m^{-β/(2β+d)})$, where $m$ counts labeled source samples and logarithmic factors are suppressed. The labeled-sample exponent is governed by the latent dimension $d$ rather than the ambient dimension $p$, which enters the pretraining cost. Experiments on Office-31, Office-Home, and DomainNet demonstrate clear transfer gains across diverse domain shifts, with particularly strong performance when representations are learned from scratch.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2610.15613",
  "author_total": 5,
  "authors": [
    {
      "name": "Yuling Jiao"
    },
    {
      "name": "Lican Kang"
    },
    {
      "name": "Jerry Zhijian Yang"
    },
    {
      "name": "Zhou Yu"
    },
    {
      "name": "Fusheng Zhou"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/9e8y9cq7t063k.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2610.15613",
  "fulltext_url": null,
  "id": "9e8y9cq7t063k",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2610.15613",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "stat.ML"
    ],
    [
      "openalex",
      "1702|Domain Adaptation and Few-Shot Learning"
    ]
  ],
  "openalex": "W7221828690",
  "pdf_url": "https://arxiv.org/pdf/2610.15613",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-11",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/9e8y9cq7t063k.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:29:26Z",
      "native_id": "2610.15613",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2610.15613"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.ML",
      "subfield": "statistics"
    }
  ],
  "title": "Joint Contrastive-Adversarial Representation Learning for Unsupervised Domain Adaptation",
  "title_text": "Joint Contrastive-Adversarial Representation Learning for Unsupervised Domain Adaptation",
  "type": "preprint",
  "updated_at": "2026-10-30T12:29:26Z",
  "url": "https://research.draconitic.com/paper/9e8y9cq7t063k",
  "venue": "arXiv",
  "versions": []
}