{"id":"W2237788583","doi":"","title":"Stepwise Variable Selection for Loglinear Mixture in Record Linkage","year":2010,"lang":"en","type":"article","venue":"","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of New Brunswick; McMaster University","funders":"","keywords":"Log-linear model; Overfitting; Mathematics; Model selection; Covariate; Selection (genetic algorithm); Mixture model; Statistical model; Linkage (software); Probabilistic logic; Statistics; Linear model; Artificial intelligence; Computer science","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":[0.02042839,0.002074949,0.003476284,0.004698406,0.00252181,0.003240173,0.007022035,0.002107921,0.005458263],"category_scores_gemma":[0.03058545,0.001640166,0.005380131,0.00522432,0.001641341,0.003799394,0.006485777,0.003502598,0.002111745],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001288008,"about_ca_system_score_gemma":0.004993141,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007718056,"about_ca_topic_score_gemma":0.008057665,"domain_scores_codex":[0.9859719,0.009429627,0.0005491903,0.002196244,0.00121502,0.000638149],"domain_scores_gemma":[0.9843975,0.01143885,0.001011085,0.001240744,0.001534106,0.0003776844],"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":[0.0005871627,0.0004321124,0.02189031,0.0005521692,0.001161181,0.0008728357,0.001507414,0.4360749,0.003060894,0.1272585,0.007226432,0.3993761],"study_design_scores_gemma":[0.00003859272,0.00007529882,0.0007836603,0.00003391796,0.00009724987,0.0001071599,0.00006374093,0.9662791,0.0007899452,0.02960359,0.00208156,0.00004623086],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00360125,0.00009453351,0.9953651,0.00009365027,0.00001784274,0.00006953706,0.00007729675,0.0005394384,0.0001414197],"genre_scores_gemma":[0.1173496,0.0003866951,0.8763897,0.0002313652,0.0001159211,0.001228068,0.001588646,0.0003840636,0.002325978],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02042839,"threshold_uncertainty_score":0.1080369,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09442882874722106,"score_gpt":0.404395673341274,"score_spread":0.309966844594053,"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."}}