{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:ZYRURGMCJUXG4I77LANFKHLI5B","short_pith_number":"pith:ZYRURGMC","schema_version":"1.0","canonical_sha256":"ce234899824d2e6e23ff581a551d68e8790fbddbbb15c911bab821949cc5bd65","source":{"kind":"arxiv","id":"2103.08933","version":1},"attestation_state":"computed","paper":{"title":"Reweighting Augmented Samples by Minimizing the Maximal Expected Loss","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lifeng Shang, Lu Hou, Mingyang Yi, Qun Liu, Xin Jiang, Zhi-Ming Ma","submitted_at":"2021-03-16T09:31:04Z","abstract_excerpt":"Data augmentation is an effective technique to improve the generalization of deep neural networks. However, previous data augmentation methods usually treat the augmented samples equally without considering their individual impacts on the model. To address this, for the augmented samples from the same training example, we propose to assign different weights to them. We construct the maximal expected loss which is the supremum over any reweighted loss on augmented samples. Inspired by adversarial training, we minimize this maximal expected loss (MMEL) and obtain a simple and interpretable close"},"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":"2103.08933","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-03-16T09:31:04Z","cross_cats_sorted":[],"title_canon_sha256":"999d170656f3d5d598b61a9811940e4d6a38ad3d8b28859e7e173c73055afbcc","abstract_canon_sha256":"f25b58a4eaa4a7c40fb70a3dff58de772d9bbd9bf5a8966812af234873ff8c49"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:23:33.300305Z","signature_b64":"m3Y3dEA7LI+NsCHnEiU+3V4FB0kz6N+DdprDiclJVgdxp1n3/ewFFdnHMBHIkgCqQgNRqfpgb12K8wWAIMiRBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ce234899824d2e6e23ff581a551d68e8790fbddbbb15c911bab821949cc5bd65","last_reissued_at":"2026-07-05T02:23:33.299898Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:23:33.299898Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reweighting Augmented Samples by Minimizing the Maximal Expected Loss","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lifeng Shang, Lu Hou, Mingyang Yi, Qun Liu, Xin Jiang, Zhi-Ming Ma","submitted_at":"2021-03-16T09:31:04Z","abstract_excerpt":"Data augmentation is an effective technique to improve the generalization of deep neural networks. However, previous data augmentation methods usually treat the augmented samples equally without considering their individual impacts on the model. To address this, for the augmented samples from the same training example, we propose to assign different weights to them. We construct the maximal expected loss which is the supremum over any reweighted loss on augmented samples. Inspired by adversarial training, we minimize this maximal expected loss (MMEL) and obtain a simple and interpretable close"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2103.08933","kind":"arxiv","version":1},"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/2103.08933/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":"2103.08933","created_at":"2026-07-05T02:23:33.299960+00:00"},{"alias_kind":"arxiv_version","alias_value":"2103.08933v1","created_at":"2026-07-05T02:23:33.299960+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2103.08933","created_at":"2026-07-05T02:23:33.299960+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZYRURGMCJUXG","created_at":"2026-07-05T02:23:33.299960+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZYRURGMCJUXG4I77","created_at":"2026-07-05T02:23:33.299960+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZYRURGMC","created_at":"2026-07-05T02:23:33.299960+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.06733","citing_title":"Dynamic Loss-Based Sample Reweighting for Improved Large Language Model Pretraining","ref_index":43,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B","json":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B.json","graph_json":"https://pith.science/api/pith-number/ZYRURGMCJUXG4I77LANFKHLI5B/graph.json","events_json":"https://pith.science/api/pith-number/ZYRURGMCJUXG4I77LANFKHLI5B/events.json","paper":"https://pith.science/paper/ZYRURGMC"},"agent_actions":{"view_html":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B","download_json":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B.json","view_paper":"https://pith.science/paper/ZYRURGMC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2103.08933&json=true","fetch_graph":"https://pith.science/api/pith-number/ZYRURGMCJUXG4I77LANFKHLI5B/graph.json","fetch_events":"https://pith.science/api/pith-number/ZYRURGMCJUXG4I77LANFKHLI5B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B/action/storage_attestation","attest_author":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B/action/author_attestation","sign_citation":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B/action/citation_signature","submit_replication":"https://pith.science/pith/ZYRURGMCJUXG4I77LANFKHLI5B/action/replication_record"}},"created_at":"2026-07-05T02:23:33.299960+00:00","updated_at":"2026-07-05T02:23:33.299960+00:00"}