{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:TPGMAGIUCT73UE2AMJOLBR3FEY","short_pith_number":"pith:TPGMAGIU","schema_version":"1.0","canonical_sha256":"9bccc0191414ffba1340625cb0c765260c6ad4548e9f2ad20b3f494f844fb031","source":{"kind":"arxiv","id":"1907.10905","version":4},"attestation_state":"computed","paper":{"title":"A Group-Theoretic Framework for Data Augmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Edgar Dobriban, Jane H Lee, Shuxiao Chen","submitted_at":"2019-07-25T08:58:59Z","abstract_excerpt":"Data augmentation is a widely used trick when training deep neural networks: in addition to the original data, properly transformed data are also added to the training set. However, to the best of our knowledge, a clear mathematical framework to explain the performance benefits of data augmentation is not available. In this paper, we develop such a theoretical framework. We show data augmentation is equivalent to an averaging operation over the orbits of a certain group that keeps the data distribution approximately invariant. We prove that it leads to variance reduction. We study empirical ri"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"1907.10905","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2019-07-25T08:58:59Z","cross_cats_sorted":["cs.LG","math.ST","stat.TH"],"title_canon_sha256":"0618be7ba25a0e0b81e469c86eda26d01f34c0bc3bf0992f96145619c7386250","abstract_canon_sha256":"0c5d5ab8831f82a7afc8d6c8fee30a323908d7c22f7b8a09312e0274639a9bb6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:49:47.313596Z","signature_b64":"xN/3Vu+VKPe1YaOEQAYAuvX0m9/RVJVhkaQMZ0JUTQJEsHCNSZyn5RTgj8zrtoTCl1ZVIcCJoC3tLEg5Ra2RCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9bccc0191414ffba1340625cb0c765260c6ad4548e9f2ad20b3f494f844fb031","last_reissued_at":"2026-07-05T01:49:47.313169Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:49:47.313169Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Group-Theoretic Framework for Data Augmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.ST","stat.TH"],"primary_cat":"stat.ML","authors_text":"Edgar Dobriban, Jane H Lee, Shuxiao Chen","submitted_at":"2019-07-25T08:58:59Z","abstract_excerpt":"Data augmentation is a widely used trick when training deep neural networks: in addition to the original data, properly transformed data are also added to the training set. However, to the best of our knowledge, a clear mathematical framework to explain the performance benefits of data augmentation is not available. In this paper, we develop such a theoretical framework. We show data augmentation is equivalent to an averaging operation over the orbits of a certain group that keeps the data distribution approximately invariant. We prove that it leads to variance reduction. We study empirical ri"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1907.10905","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1907.10905/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"1907.10905","created_at":"2026-07-05T01:49:47.313222+00:00"},{"alias_kind":"arxiv_version","alias_value":"1907.10905v4","created_at":"2026-07-05T01:49:47.313222+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1907.10905","created_at":"2026-07-05T01:49:47.313222+00:00"},{"alias_kind":"pith_short_12","alias_value":"TPGMAGIUCT73","created_at":"2026-07-05T01:49:47.313222+00:00"},{"alias_kind":"pith_short_16","alias_value":"TPGMAGIUCT73UE2A","created_at":"2026-07-05T01:49:47.313222+00:00"},{"alias_kind":"pith_short_8","alias_value":"TPGMAGIU","created_at":"2026-07-05T01:49:47.313222+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY","json":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY.json","graph_json":"https://pith.science/api/pith-number/TPGMAGIUCT73UE2AMJOLBR3FEY/graph.json","events_json":"https://pith.science/api/pith-number/TPGMAGIUCT73UE2AMJOLBR3FEY/events.json","paper":"https://pith.science/paper/TPGMAGIU"},"agent_actions":{"view_html":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY","download_json":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY.json","view_paper":"https://pith.science/paper/TPGMAGIU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1907.10905&json=true","fetch_graph":"https://pith.science/api/pith-number/TPGMAGIUCT73UE2AMJOLBR3FEY/graph.json","fetch_events":"https://pith.science/api/pith-number/TPGMAGIUCT73UE2AMJOLBR3FEY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY/action/storage_attestation","attest_author":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY/action/author_attestation","sign_citation":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY/action/citation_signature","submit_replication":"https://pith.science/pith/TPGMAGIUCT73UE2AMJOLBR3FEY/action/replication_record"}},"created_at":"2026-07-05T01:49:47.313222+00:00","updated_at":"2026-07-05T01:49:47.313222+00:00"}