{"id":"W1999376700","doi":"10.1016/j.tcs.2012.10.013","title":"Tighter PAC-Bayes bounds through distribution-dependent priors","year":2012,"lang":"en","type":"article","venue":"Theoretical Computer Science","topic":"Machine Learning and Algorithms","field":"Computer Science","cited_by":75,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université Laval","funders":"","keywords":"Prior probability; Bayes' theorem; Mathematics; Reproducing kernel Hilbert space; Regularization (linguistics); Mathematical optimization; Applied mathematics; Exponential family; Algorithm; Class (philosophy); Function (biology); Distribution (mathematics); Computer science; Hilbert space; Artificial intelligence; Bayesian probability; Statistics; Mathematical analysis","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02442865,0.004749354,0.005889921,0.003936542,0.00308106,0.008727347,0.006537524,0.005706833,0.01590188],"category_scores_gemma":[0.1343143,0.004338477,0.002848298,0.004736926,0.007082033,0.02396195,0.01111773,0.01905622,0.004234355],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006213488,"about_ca_system_score_gemma":0.005962928,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004283873,"about_ca_topic_score_gemma":0.007330006,"domain_scores_codex":[0.9804388,0.008498205,0.0007061744,0.002843843,0.005839943,0.001673024],"domain_scores_gemma":[0.8885446,0.09400424,0.002146596,0.009916694,0.003787975,0.00159982],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0005863367,0.0002329041,0.001248819,0.000487281,0.0002028034,0.0002079119,0.0003703079,0.1478463,0.001267381,0.7673877,0.01787676,0.06228544],"study_design_scores_gemma":[0.00004821291,0.00002558289,0.00017641,0.0001031954,0.00004538459,0.00008545606,0.00003064179,0.2086903,0.0005522135,0.7875804,0.002632376,0.00002997053],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008352932,0.005144183,0.9587742,0.00471911,0.0005491191,0.00008820871,0.0004673567,0.000852181,0.02105268],"genre_scores_gemma":[0.4263748,0.008062733,0.5233052,0.00744865,0.003554808,0.0009455394,0.002078983,0.002656586,0.02557284],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02442865,"threshold_uncertainty_score":0.1291926,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.007516441666955032,"score_gpt":0.2596614497271334,"score_spread":0.2521450080601784,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}