{"id":"W4389372502","doi":"10.1136/archdischild-2023-326340","title":"Building a better model: abandon kitchen sink regression","year":2023,"lang":"en","type":"review","venue":"Archives of Disease in Childhood Fetal & Neonatal","topic":"Advanced Causal Inference Techniques","field":"Mathematics","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of New Brunswick; Dalhousie University","funders":"","keywords":"Overfitting; Directed acyclic graph; Regression; Regression analysis; Computer science; Outcome (game theory); Feature selection; Model selection; Econometrics; Logistic regression; Machine learning; Artificial intelligence; Statistics; Mathematics","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.02299861,0.001637604,0.003345488,0.00269295,0.0004267014,0.002952346,0.003706492,0.002335986,0.004218157],"category_scores_gemma":[0.06154685,0.0008556838,0.003075307,0.002279266,0.00270788,0.005523161,0.002092705,0.007924922,0.002387102],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002031265,"about_ca_system_score_gemma":0.003639108,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003724804,"about_ca_topic_score_gemma":0.003716435,"domain_scores_codex":[0.9874601,0.008910619,0.0006974555,0.001180138,0.00161709,0.0001346076],"domain_scores_gemma":[0.9459052,0.04882486,0.001365631,0.001942885,0.001726097,0.0002352828],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001123154,0.00006114391,0.001521314,0.009896432,0.001516583,0.0002745006,0.0004875392,0.04002027,0.0003505727,0.3877469,0.03084649,0.5271659],"study_design_scores_gemma":[0.00008186646,0.0002233787,0.0007575559,0.006692166,0.0005202186,0.0005334074,0.0002671595,0.05453074,0.0006629262,0.6217676,0.3138302,0.000132888],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"review","genre_scores_codex":[0.001561663,0.2675329,0.6999195,0.02448125,0.001608694,0.000102187,0.0003614786,0.0005307956,0.003901626],"genre_scores_gemma":[0.07208068,0.3452186,0.5568851,0.01388043,0.004018401,0.0006482588,0.0008549488,0.0009148194,0.005498762],"genre_candidate":"review","genre_consensus":null,"teacher_disagreement_score":0.9770014,"threshold_uncertainty_score":0.1216298,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.103280087541211,"score_gpt":0.4137124541359426,"score_spread":0.3104323665947316,"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."}}