{"paper":{"title":"Optimal Spatio-Temporal Decoupling for Bayesian Conformal Prediction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"State-Adaptive Bayesian Conformal Prediction gates temporal inertia with spatial kernel-density evidence to balance coverage and efficiency.","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chia-Yen Lee, Yu-Hsueh Fang","submitted_at":"2026-05-01T06:07:07Z","abstract_excerpt":"Online conformal prediction must balance fast adaptation to distribution shift against stable coverage: feedback-driven methods react quickly but become volatile, while strongly discounted Bayesian methods lag and inflate intervals at tight coverage. We introduce \\textbf{State-Adaptive Bayesian Conformal Prediction (SA-BCP)}, which forms the predictive quantile as a gated convex combination of long-term temporal inertia and local spatial evidence from a kernel density estimate, controlled by a single interpretable evidence threshold $K$. We establish three results: (i) asymptotic marginal vali"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We rigorously prove this mechanism's optimality, identifying a minimax bias-variance tradeoff governed by an evidence threshold K. ... SA-BCP resolves the systematic under-coverage inherent to ACI variants while simultaneously reducing the uncalibrated interval bloat of Bayesian CP by 10% to 37% under high-confidence requests.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That spatial kernel-density evidence can accurately and proactively identify historical regimes to gate temporal inertia without introducing new lag or calibration errors, as stated in the description of the SA-BCP mechanism.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SA-BCP achieves optimal spatio-temporal decoupling in Bayesian conformal prediction by gating temporal inertia with spatial kernel-density evidence, governed by a minimax bias-variance threshold K, and outperforms ACI and Bayesian CP baselines on financial datasets.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"State-Adaptive Bayesian Conformal Prediction gates temporal inertia with spatial kernel-density evidence to balance coverage and efficiency.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"445524a2021d9de967b6a8c4e4d5078f385c2b2a135299caddb8b666fe93a976"},"source":{"id":"2605.00432","kind":"arxiv","version":2},"verdict":{"id":"2016952b-8668-4996-b256-364df879dbc5","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T19:45:51.887611Z","strongest_claim":"We rigorously prove this mechanism's optimality, identifying a minimax bias-variance tradeoff governed by an evidence threshold K. ... SA-BCP resolves the systematic under-coverage inherent to ACI variants while simultaneously reducing the uncalibrated interval bloat of Bayesian CP by 10% to 37% under high-confidence requests.","one_line_summary":"SA-BCP achieves optimal spatio-temporal decoupling in Bayesian conformal prediction by gating temporal inertia with spatial kernel-density evidence, governed by a minimax bias-variance threshold K, and outperforms ACI and Bayesian CP baselines on financial datasets.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That spatial kernel-density evidence can accurately and proactively identify historical regimes to gate temporal inertia without introducing new lag or calibration errors, as stated in the description of the SA-BCP mechanism.","pith_extraction_headline":"State-Adaptive Bayesian Conformal Prediction gates temporal inertia with spatial kernel-density evidence to balance coverage and efficiency."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.00432/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T19:42:51.420150Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T18:10:35.961548Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"778bd9108d0ae8574c77005d93cef725cee057891d3211270518df5a0ede8478"},"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"}