{
  "annotation_id": "e058b90a-c69b-4e70-85e9-dbdd88571527",
  "date_created": "2026-03-02T18:01:00.856000Z",
  "date_modified": "2026-03-02T18:01:00.856000Z",
  "file_hash": "2a38afb10e55ae810759f00e75eb95930d9e6a772373534365775454fb60721a",
  "private": false,
  "record": {
    "abstract": "In the framework of on-line learning, a learning machine might move around a\nteacher due to the differences in structures or output functions between the\nteacher and the learning machine or due to noises. The generalization\nperformance of a new student supervised by a moving machine has been analyzed.\nA model composed of a true teacher, a moving teacher and a student that are all\nlinear perceptrons with noises has been treated analytically using statistical\nmechanics. It has been proven that the generalization errors of a student can\nbe smaller than that of a moving teacher, even if the student only uses\nexamples from the moving teacher.",
    "arxiv_id": "physics/0509050",
    "authors": [
      "Seiji Miyoshi",
      "Masato Okada"
    ],
    "categories": [
      "physics.soc-ph",
      "cond-mat.dis-nn"
    ],
    "doi": "10.1143/JPSJ.75.024003",
    "title": "Analysis of on-line learning when a moving teacher goes around a true   teacher",
    "url": "https://arxiv.org/abs/physics/0509050"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "c2332649-35c0-46bd-b355-b91dd8e8a83d",
    "id": "arXiv Dataset IDs",
    "type": "Model",
    "variant": "snapshot-2026-03-01",
    "version": "0.1.0"
  },
  "user_id": 1000002
}