{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7XVDKBISHLGOFDTDRKDLIY2LSC","short_pith_number":"pith:7XVDKBIS","schema_version":"1.0","canonical_sha256":"fdea3505123acce28e638a86b4634b90a3edac010a92c4a79569157eac3e9eac","source":{"kind":"arxiv","id":"2603.29981","version":3},"attestation_state":"computed","paper":{"title":"Aligning Validation with Deployment in Spatial Prediction: Target-Weighted Cross-Validation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Target-weighted cross-validation reduces bias in spatial prediction performance estimates by aligning validation tasks with the full deployment domain.","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexander Brenning, Thomas Suesse","submitted_at":"2026-03-31T16:44:07Z","abstract_excerpt":"Reliable estimation of predictive performance is essential for spatial environmental modeling, where machine-learning models are used to generate maps from unevenly distributed observations. Standard cross-validation (CV) assumes that validation data are representative of prediction conditions across the target domain. In practice, this assumption is often violated due to preferential or clustered sampling, leading to biased performance and uncertainty estimates.\n  We introduce a deployment-oriented validation framework based on weighted CV that aligns validation tasks with the distribution of"},"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":true},"canonical_record":{"source":{"id":"2603.29981","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-31T16:44:07Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"e68a6672ad0582c2858bcd7591ffbcd95748cdc7cd44887b2ff602f0527f6590","abstract_canon_sha256":"b2b2b3850c5aca337c5c0cc1a2545ddcd479b822da4a8085bcdd4287a97896c2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T02:04:40.600574Z","signature_b64":"ddfXZQORWKBYZMjvkoRox1NH7xDNALD0/LbAaR1tcWdQu/A7IZU+BANWJKwUDETA6aEj2BDhPVLX1ymLdGhDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fdea3505123acce28e638a86b4634b90a3edac010a92c4a79569157eac3e9eac","last_reissued_at":"2026-05-22T02:04:40.599562Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T02:04:40.599562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Aligning Validation with Deployment in Spatial Prediction: Target-Weighted Cross-Validation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Target-weighted cross-validation reduces bias in spatial prediction performance estimates by aligning validation tasks with the full deployment domain.","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexander Brenning, Thomas Suesse","submitted_at":"2026-03-31T16:44:07Z","abstract_excerpt":"Reliable estimation of predictive performance is essential for spatial environmental modeling, where machine-learning models are used to generate maps from unevenly distributed observations. Standard cross-validation (CV) assumes that validation data are representative of prediction conditions across the target domain. In practice, this assumption is often violated due to preferential or clustered sampling, leading to biased performance and uncertainty estimates.\n  We introduce a deployment-oriented validation framework based on weighted CV that aligns validation tasks with the distribution of"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"weighted CV approaches substantially reduce this bias when validation tasks adequately cover the deployment-task space.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That spatially meaningful task descriptors such as environmental covariates and prediction distance are sufficient to represent and cover the full deployment-task space for weighting purposes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Target-Weighted Cross-Validation reduces bias in spatial model performance estimates by weighting validation tasks to match the distribution of prediction tasks across the deployment domain.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Target-weighted cross-validation reduces bias in spatial prediction performance estimates by aligning validation tasks with the full deployment domain.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c38641dcc0c532b6e58b9e5bf85ea1db9f83120a95efdc544959cc7f3e542841"},"source":{"id":"2603.29981","kind":"arxiv","version":3},"verdict":{"id":"d50fab23-d463-4d0a-830d-7a3376a8d895","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T23:37:18.093655Z","strongest_claim":"weighted CV approaches substantially reduce this bias when validation tasks adequately cover the deployment-task space.","one_line_summary":"Target-Weighted Cross-Validation reduces bias in spatial model performance estimates by weighting validation tasks to match the distribution of prediction tasks across the deployment domain.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That spatially meaningful task descriptors such as environmental covariates and prediction distance are sufficient to represent and cover the full deployment-task space for weighting purposes.","pith_extraction_headline":"Target-weighted cross-validation reduces bias in spatial prediction performance estimates by aligning validation tasks with the full deployment domain."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.29981/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":2,"snapshot_sha256":"7b076251a842f673bd61931e8f1a16694afeb56b70dc37a86c96f68b41816e2c"},"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":"2603.29981","created_at":"2026-05-22T02:04:40.599707+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.29981v3","created_at":"2026-05-22T02:04:40.599707+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.29981","created_at":"2026-05-22T02:04:40.599707+00:00"},{"alias_kind":"pith_short_12","alias_value":"7XVDKBISHLGO","created_at":"2026-05-22T02:04:40.599707+00:00"},{"alias_kind":"pith_short_16","alias_value":"7XVDKBISHLGOFDTD","created_at":"2026-05-22T02:04:40.599707+00:00"},{"alias_kind":"pith_short_8","alias_value":"7XVDKBIS","created_at":"2026-05-22T02:04:40.599707+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.13689","citing_title":"Moving beyond spatial and random cross-validation in environmental modelling: a call for prediction-domain adaptive evaluation","ref_index":44,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC","json":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC.json","graph_json":"https://pith.science/api/pith-number/7XVDKBISHLGOFDTDRKDLIY2LSC/graph.json","events_json":"https://pith.science/api/pith-number/7XVDKBISHLGOFDTDRKDLIY2LSC/events.json","paper":"https://pith.science/paper/7XVDKBIS"},"agent_actions":{"view_html":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC","download_json":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC.json","view_paper":"https://pith.science/paper/7XVDKBIS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.29981&json=true","fetch_graph":"https://pith.science/api/pith-number/7XVDKBISHLGOFDTDRKDLIY2LSC/graph.json","fetch_events":"https://pith.science/api/pith-number/7XVDKBISHLGOFDTDRKDLIY2LSC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC/action/storage_attestation","attest_author":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC/action/author_attestation","sign_citation":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC/action/citation_signature","submit_replication":"https://pith.science/pith/7XVDKBISHLGOFDTDRKDLIY2LSC/action/replication_record"}},"created_at":"2026-05-22T02:04:40.599707+00:00","updated_at":"2026-05-22T02:04:40.599707+00:00"}