{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:F6LJLNEWR2DHLTHNJAQYMR33E3","short_pith_number":"pith:F6LJLNEW","schema_version":"1.0","canonical_sha256":"2f9695b4968e8675cced482186477b26e56e0a224bcb13dfdcc264ef2bc99a58","source":{"kind":"arxiv","id":"2503.00877","version":2},"attestation_state":"computed","paper":{"title":"Patch-wise Structural Loss for Time Series Forecasting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dilfira Kudrat, Qinghua Hu, Tianyu Jia, Yanru Sun, Zongxia Xie","submitted_at":"2025-03-02T12:36:15Z","abstract_excerpt":"Time-series forecasting has gained significant attention in machine learning due to its crucial role in various domains. However, most existing forecasting models rely heavily on point-wise loss functions like Mean Square Error, which treat each time step independently and neglect the structural dependencies inherent in time series data, making it challenging to capture complex temporal patterns accurately. To address these challenges, we propose a novel Patch-wise Structural (PS) loss, designed to enhance structural alignment by comparing time series at the patch level. Through leveraging loc"},"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":"2503.00877","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-02T12:36:15Z","cross_cats_sorted":[],"title_canon_sha256":"443f3d07e65d1ce49c5e5d3aa2e96fbed6250a9872139d69c7ec0ac62c4d4700","abstract_canon_sha256":"8e963f1c5c13dd61428f2ee490b1eb505983c17c8822d318da0b66fdca8d4c8a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:16.576600Z","signature_b64":"8gEXQ5os3aaUChayDGJtLH0fpIsL6vtvGiyi98L0PzbgORNgc0LdLrjfXmydZ0gafuCPTFnCrB1NOjNOhCnzBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f9695b4968e8675cced482186477b26e56e0a224bcb13dfdcc264ef2bc99a58","last_reissued_at":"2026-07-05T11:37:16.576102Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:16.576102Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Patch-wise Structural Loss for Time Series Forecasting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dilfira Kudrat, Qinghua Hu, Tianyu Jia, Yanru Sun, Zongxia Xie","submitted_at":"2025-03-02T12:36:15Z","abstract_excerpt":"Time-series forecasting has gained significant attention in machine learning due to its crucial role in various domains. However, most existing forecasting models rely heavily on point-wise loss functions like Mean Square Error, which treat each time step independently and neglect the structural dependencies inherent in time series data, making it challenging to capture complex temporal patterns accurately. To address these challenges, we propose a novel Patch-wise Structural (PS) loss, designed to enhance structural alignment by comparing time series at the patch level. Through leveraging loc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.00877","kind":"arxiv","version":2},"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/2503.00877/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":"2503.00877","created_at":"2026-07-05T11:37:16.576161+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.00877v2","created_at":"2026-07-05T11:37:16.576161+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.00877","created_at":"2026-07-05T11:37:16.576161+00:00"},{"alias_kind":"pith_short_12","alias_value":"F6LJLNEWR2DH","created_at":"2026-07-05T11:37:16.576161+00:00"},{"alias_kind":"pith_short_16","alias_value":"F6LJLNEWR2DHLTHN","created_at":"2026-07-05T11:37:16.576161+00:00"},{"alias_kind":"pith_short_8","alias_value":"F6LJLNEW","created_at":"2026-07-05T11:37:16.576161+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25439","citing_title":"TopoCast: A Topological Fidelity Framework for Evaluating Transformer-Based Time Series Forecasting","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21072","citing_title":"SqLinear: Balanced Square Partitioning Makes Linear Interaction Sufficient for Large-Scale Traffic Forecasting","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02138","citing_title":"VLBM: Variational Latent Basis Modeling for OOD Robust Multivariate Time Series Forecasting","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25943","citing_title":"STaT: Resolving Shape Distortion in Non-Stationary Time Series via Tri-Modal Synergy","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2509.25914","citing_title":"ReNF: Rethinking the Design of Neural Long-Term Time Series Forecasters","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2602.00297","citing_title":"From Observations to States: Latent Time Series Forecasting","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3","json":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3.json","graph_json":"https://pith.science/api/pith-number/F6LJLNEWR2DHLTHNJAQYMR33E3/graph.json","events_json":"https://pith.science/api/pith-number/F6LJLNEWR2DHLTHNJAQYMR33E3/events.json","paper":"https://pith.science/paper/F6LJLNEW"},"agent_actions":{"view_html":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3","download_json":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3.json","view_paper":"https://pith.science/paper/F6LJLNEW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.00877&json=true","fetch_graph":"https://pith.science/api/pith-number/F6LJLNEWR2DHLTHNJAQYMR33E3/graph.json","fetch_events":"https://pith.science/api/pith-number/F6LJLNEWR2DHLTHNJAQYMR33E3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3/action/storage_attestation","attest_author":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3/action/author_attestation","sign_citation":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3/action/citation_signature","submit_replication":"https://pith.science/pith/F6LJLNEWR2DHLTHNJAQYMR33E3/action/replication_record"}},"created_at":"2026-07-05T11:37:16.576161+00:00","updated_at":"2026-07-05T11:37:16.576161+00:00"}