{
  "id": "2206.08476",
  "version": "v1",
  "published": "2022-06-16T22:52:08.000Z",
  "updated": "2022-06-16T22:52:08.000Z",
  "title": "Zero-Shot AutoML with Pretrained Models",
  "authors": [
    "Ekrem Öztürk",
    "Fabio Ferreira",
    "Hadi S. Jomaa",
    "Lars Schmidt-Thieme",
    "Josif Grabocka",
    "Frank Hutter"
  ],
  "journal": "International Conference on Machine Learning 2022",
  "categories": [
    "cs.LG",
    "cs.AI",
    "cs.CV"
  ],
  "abstract": "Given a new dataset D and a low compute budget, how should we choose a pre-trained model to fine-tune to D, and set the fine-tuning hyperparameters without risking overfitting, particularly if D is small? Here, we extend automated machine learning (AutoML) to best make these choices. Our domain-independent meta-learning approach learns a zero-shot surrogate model which, at test time, allows to select the right deep learning (DL) pipeline (including the pre-trained model and fine-tuning hyperparameters) for a new dataset D given only trivial meta-features describing D such as image resolution or the number of classes. To train this zero-shot model, we collect performance data for many DL pipelines on a large collection of datasets and meta-train on this data to minimize a pairwise ranking objective. We evaluate our approach under the strict time limit of the vision track of the ChaLearn AutoDL challenge benchmark, clearly outperforming all challenge contenders.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2022-06-16T22:52:08.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "zero-shot automl",
      "pretrained models",
      "chalearn autodl challenge benchmark",
      "pre-trained model",
      "strict time limit"
    ],
    "tags": [
      "journal article"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}