{"id":"W4402406483","doi":"10.23889/ijpds.v9i5.2846","title":"Model-based algorithms to ascertain smoking in administrative health data: a registry-based validation study","year":2024,"lang":"en","type":"article","venue":"International Journal for Population Data Science","topic":"demographic modeling and climate adaptation","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Manitoba","funders":"","keywords":"Computer science; Data mining; Algorithm; Data science","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.06129435,0.001339973,0.001188541,0.002120731,0.0006002628,0.00165177,0.002327589,0.001329403,0.0008670482],"category_scores_gemma":[0.1179107,0.0005613533,0.001945434,0.001416467,0.0005565176,0.001350112,0.001536817,0.001968575,0.0003472582],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001723001,"about_ca_system_score_gemma":0.003201036,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01348437,"about_ca_topic_score_gemma":0.007048775,"domain_scores_codex":[0.9753975,0.01887297,0.001382421,0.001974507,0.001892014,0.0004805485],"domain_scores_gemma":[0.9026707,0.07776303,0.003712216,0.006348488,0.008831092,0.0006744764],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001888484,0.003169252,0.5828032,0.000449819,0.003236281,0.000182871,0.0003176941,0.2545781,0.000776753,0.002518564,0.004115694,0.1459634],"study_design_scores_gemma":[0.0005492721,0.001068974,0.04333638,0.0001263142,0.0003808615,0.000253344,0.00008425858,0.9504634,0.0007022647,0.001861656,0.001128598,0.0000448315],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.8294616,0.001640819,0.1612295,0.000676601,0.0001158341,0.001455255,0.002861511,0.0009674547,0.001591577],"genre_scores_gemma":[0.8802617,0.0003464293,0.1147469,0.0002154964,0.00003954745,0.0007802238,0.003301063,0.00008889711,0.0002197442],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.06129435,"threshold_uncertainty_score":0.3241594,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5534976125645847,"score_gpt":0.586329769139679,"score_spread":0.03283215657509431,"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."}}