{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:2DCJKKME35E6ZEH5ZJCAID2TKU","short_pith_number":"pith:2DCJKKME","schema_version":"1.0","canonical_sha256":"d0c4952984df49ec90fdca44040f535526b96d0529af2ed9a0e025a184e66c98","source":{"kind":"arxiv","id":"1908.00690","version":1},"attestation_state":"computed","paper":{"title":"Feature Robustness in Non-stationary Health Records: Caveats to Deployable Model Performance in Common Clinical Machine Learning Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anna Goldenberg, Bret Nestor, Gabriela Berner, Marzyeh Ghassemi, Matthew B. A. McDermott, Michael C. Hughes, Tristan Naumann, Willie Boag","submitted_at":"2019-08-02T03:03:25Z","abstract_excerpt":"When training clinical prediction models from electronic health records (EHRs), a key concern should be a model's ability to sustain performance over time when deployed, even as care practices, database systems, and population demographics evolve. Due to de-identification requirements, however, current experimental practices for public EHR benchmarks (such as the MIMIC-III critical care dataset) are time agnostic, assigning care records to train or test sets without regard for the actual dates of care. As a result, current benchmarks cannot assess how well models trained on one year generalise"},"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":"1908.00690","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2019-08-02T03:03:25Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"b63d6edb7f486399daf3be88d020309f0cf5e433d9f80d73afc381366f01e069","abstract_canon_sha256":"ae40aed790087b21ef509f534df55b0627e156c6597358968ea2e54cc1301d02"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-04T23:51:08.445137Z","signature_b64":"hXICJzjTY87logkvJgQja5MKBAEcCBIjJr9UdIVlu9gyjBMoMGImmOKagOHzt+YyjeY8qPKohwOlR/XCpjYXDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0c4952984df49ec90fdca44040f535526b96d0529af2ed9a0e025a184e66c98","last_reissued_at":"2026-07-04T23:51:08.444552Z","signature_status":"signed_v1","first_computed_at":"2026-07-04T23:51:08.444552Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Feature Robustness in Non-stationary Health Records: Caveats to Deployable Model Performance in Common Clinical Machine Learning Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anna Goldenberg, Bret Nestor, Gabriela Berner, Marzyeh Ghassemi, Matthew B. A. McDermott, Michael C. Hughes, Tristan Naumann, Willie Boag","submitted_at":"2019-08-02T03:03:25Z","abstract_excerpt":"When training clinical prediction models from electronic health records (EHRs), a key concern should be a model's ability to sustain performance over time when deployed, even as care practices, database systems, and population demographics evolve. Due to de-identification requirements, however, current experimental practices for public EHR benchmarks (such as the MIMIC-III critical care dataset) are time agnostic, assigning care records to train or test sets without regard for the actual dates of care. As a result, current benchmarks cannot assess how well models trained on one year generalise"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.00690","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/1908.00690/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":"1908.00690","created_at":"2026-07-04T23:51:08.444621+00:00"},{"alias_kind":"arxiv_version","alias_value":"1908.00690v1","created_at":"2026-07-04T23:51:08.444621+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.00690","created_at":"2026-07-04T23:51:08.444621+00:00"},{"alias_kind":"pith_short_12","alias_value":"2DCJKKME35E6","created_at":"2026-07-04T23:51:08.444621+00:00"},{"alias_kind":"pith_short_16","alias_value":"2DCJKKME35E6ZEH5","created_at":"2026-07-04T23:51:08.444621+00:00"},{"alias_kind":"pith_short_8","alias_value":"2DCJKKME","created_at":"2026-07-04T23:51:08.444621+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/2DCJKKME35E6ZEH5ZJCAID2TKU","json":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU.json","graph_json":"https://pith.science/api/pith-number/2DCJKKME35E6ZEH5ZJCAID2TKU/graph.json","events_json":"https://pith.science/api/pith-number/2DCJKKME35E6ZEH5ZJCAID2TKU/events.json","paper":"https://pith.science/paper/2DCJKKME"},"agent_actions":{"view_html":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU","download_json":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU.json","view_paper":"https://pith.science/paper/2DCJKKME","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1908.00690&json=true","fetch_graph":"https://pith.science/api/pith-number/2DCJKKME35E6ZEH5ZJCAID2TKU/graph.json","fetch_events":"https://pith.science/api/pith-number/2DCJKKME35E6ZEH5ZJCAID2TKU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU/action/storage_attestation","attest_author":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU/action/author_attestation","sign_citation":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU/action/citation_signature","submit_replication":"https://pith.science/pith/2DCJKKME35E6ZEH5ZJCAID2TKU/action/replication_record"}},"created_at":"2026-07-04T23:51:08.444621+00:00","updated_at":"2026-07-04T23:51:08.444621+00:00"}