{"id":"W3173332884","doi":"10.48550/arxiv.2106.12535","title":"Learning Stochastic Majority Votes by Minimizing a PAC-Bayes Generalization Bound","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Machine Learning and Algorithms","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"","keywords":"Generalization; Prior probability; Differentiable function; Computer science; Bayes' theorem; Series (stratigraphy); Artificial intelligence; Dirichlet distribution; Upper and lower bounds; Algorithm; Mathematics; Applied mathematics; Machine learning; Mathematical optimization; Bayesian probability","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0003169922,0.0003785309,0.0003936913,0.000210081,0.0003834226,0.0006154769,0.001143503,0.0002905399,0.00004232724],"category_scores_gemma":[0.0001713812,0.0004623059,0.0002081117,0.0006350515,0.00007937779,0.0004019993,0.001673786,0.001113671,0.000024588],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001618796,"about_ca_system_score_gemma":0.0002103351,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007188436,"about_ca_topic_score_gemma":0.00003676135,"domain_scores_codex":[0.997422,0.0004213909,0.000226689,0.001332554,0.0001613496,0.0004360405],"domain_scores_gemma":[0.9983744,0.0001603318,0.0003247036,0.0007641695,0.0001923129,0.0001840664],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000006689345,0.00008986,0.002846456,0.0000949896,0.0001027453,0.0001882991,0.001021562,0.9830816,0.0002032109,0.009394942,0.0003005541,0.002669102],"study_design_scores_gemma":[0.0003160704,0.00005434482,0.0003934366,0.0001369094,0.00005656,0.000008466538,0.0001621512,0.9949644,0.0001098731,0.002722384,0.0005452318,0.0005302326],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2894997,0.0002297671,0.7089349,0.0001005756,0.0005208658,0.0001121344,0.000005202753,0.0003213939,0.0002754817],"genre_scores_gemma":[0.9910149,0.0000968172,0.005815865,0.00007653376,0.0001551701,0.000001214182,0.0001280862,0.00002730082,0.002684137],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.703119,"threshold_uncertainty_score":0.9997829,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03040209435461501,"score_gpt":0.1859151147798472,"score_spread":0.1555130204252322,"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."}}