{"id":"W2161493293","doi":"10.1109/crv.2009.30","title":"Optimal Online Data Sampling or How to Hire the Best Secretaries","year":2009,"lang":"en","type":"article","venue":"","topic":"Optimization and Search Problems","field":"Computer Science","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Generalization; Sampling (signal processing); Sample (material); Selection (genetic algorithm); Secretary problem; Simple (philosophy); Artificial intelligence; Machine learning; Online algorithm; Sequence (biology); Data mining; Mathematical optimization; Algorithm; Mathematics; Computer vision","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":[0.005829754,0.0009414491,0.0025199,0.0008053941,0.001001943,0.00212548,0.002380208,0.00247661,0.003859274],"category_scores_gemma":[0.02513713,0.0007365036,0.001183725,0.001347906,0.002948693,0.004841029,0.002215669,0.002163891,0.001035351],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009378644,"about_ca_system_score_gemma":0.00223429,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001052755,"about_ca_topic_score_gemma":0.001414928,"domain_scores_codex":[0.9961862,0.001918021,0.0002066042,0.0009402229,0.0004364277,0.0003124645],"domain_scores_gemma":[0.9811136,0.01383511,0.001190564,0.002825391,0.0004985011,0.0005369722],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002845441,0.001008255,0.009353417,0.001095114,0.0004254289,0.000326164,0.0007541393,0.3035292,0.01513961,0.1181417,0.01581347,0.5315679],"study_design_scores_gemma":[0.0002451699,0.0003818473,0.001370951,0.00007079265,0.0000923474,0.000312124,0.0002656895,0.7747253,0.01018129,0.2072912,0.004991399,0.00007196949],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02962643,0.0007478544,0.965759,0.001498362,0.00008583643,0.0001965649,0.0002321094,0.0004789074,0.001374965],"genre_scores_gemma":[0.4957374,0.0007241078,0.4970505,0.0006990951,0.0004026838,0.000483256,0.0007386375,0.0002374792,0.003926909],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005829754,"threshold_uncertainty_score":0.03083104,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1732802181241352,"score_gpt":0.3577267200153544,"score_spread":0.1844465018912191,"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."}}