{"paper":{"title":"MEC: Machine-Learning-Assisted Generalized Entropy Calibration for Semi-Supervised Mean Estimation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Machine-learning-assisted generalized entropy calibration attains the semiparametric efficiency bound for semi-supervised mean estimation under weaker assumptions than prior PPI variants.","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Jae Kwang Kim, Se Yoon Lee","submitted_at":"2026-04-07T05:30:11Z","abstract_excerpt":"Obtaining high-quality labels is costly, whereas unlabeled covariates are often abundant, motivating semi-supervised inference methods with reliable uncertainty quantification. Prediction-powered inference (PPI) leverages a machine-learning predictor trained on a small labeled sample to improve efficiency, but it can lose efficiency under model misspecification and suffer from coverage distortions due to label reuse. We introduce Machine-Learning-Assisted Generalized Entropy Calibration (MEC), a cross-fitted, calibration-weighted variant of PPI. MEC improves efficiency by reweighting labeled s"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"MEC attains the semiparametric efficiency bound under weaker assumptions than existing PPI variants.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The calibration framework based on Bregman projections produces weights that align labeled samples with the target population and that the weaker projection-error conditions suffice for validity and efficiency.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MEC attains the semiparametric efficiency bound for mean estimation under weaker projection-error conditions than prior PPI methods by using generalized entropy calibration.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Machine-learning-assisted generalized entropy calibration attains the semiparametric efficiency bound for semi-supervised mean estimation under weaker assumptions than prior PPI variants.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"317bc9966679cd489acee878cdbe4063443b8b0c654e9c809c37384718da90f1"},"source":{"id":"2604.05446","kind":"arxiv","version":2},"verdict":{"id":"b5be9364-8b06-4e8a-abc9-3016a6ff49a7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T19:29:58.890349Z","strongest_claim":"MEC attains the semiparametric efficiency bound under weaker assumptions than existing PPI variants.","one_line_summary":"MEC attains the semiparametric efficiency bound for mean estimation under weaker projection-error conditions than prior PPI methods by using generalized entropy calibration.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The calibration framework based on Bregman projections produces weights that align labeled samples with the target population and that the weaker projection-error conditions suffice for validity and efficiency.","pith_extraction_headline":"Machine-learning-assisted generalized entropy calibration attains the semiparametric efficiency bound for semi-supervised mean estimation under weaker assumptions than prior PPI variants."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.05446/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":"092a132ebccc11e0f274390f9e147d7ac23d2b1e9642034bf715dee7c4686e51"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}