{"id":"W4401273264","doi":"10.1186/s12874-024-02284-5","title":"Can supervised deep learning architecture outperform autoencoders in building propensity score models for matching?","year":2024,"lang":"en","type":"article","venue":"BMC Medical Research Methodology","topic":"Advanced Causal Inference Techniques","field":"Mathematics","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre for Advancing Health Outcomes; University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Autoencoder; Propensity score matching; Artificial intelligence; Computer science; Machine learning; Deep learning; Supervised learning; Matching (statistics); Unsupervised learning; Observational study; Variance (accounting); Statistics; Artificial neural network; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":["metaresearch","research_integrity"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.03670565,0.0002852754,0.000807647,0.0007947055,0.0002290521,0.00007876803,0.0007907074,0.0005651945,0.0002006633],"category_scores_gemma":[0.09709975,0.0002225706,0.0001598985,0.0007049383,0.0005977142,0.0001750718,0.0005622273,0.003536839,0.000005454438],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003835712,"about_ca_system_score_gemma":0.001311009,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005268015,"about_ca_topic_score_gemma":0.004681815,"domain_scores_codex":[0.9897632,0.005975339,0.0006849127,0.0008344821,0.001398482,0.001343561],"domain_scores_gemma":[0.938536,0.06029091,0.00005784327,0.000425229,0.0002639699,0.0004259767],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000516354,0.0001296365,0.0007996911,0.004294736,0.00007392099,0.0002912997,0.009550062,0.006273111,0.005440981,0.739445,0.0003024263,0.2328828],"study_design_scores_gemma":[0.0002873273,0.0002506132,0.00003449135,0.0006747518,0.000009880229,0.00004994519,0.0005641535,0.2363728,0.001360049,0.7599655,0.0002484682,0.0001820368],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1255133,0.0003384064,0.8703568,0.001558069,0.0001366724,0.001181571,0.000004614623,0.000472175,0.00043834],"genre_scores_gemma":[0.2190154,0.0001359342,0.7796817,0.0001024626,0.0001743266,0.0005362586,0.000009674103,0.00008047192,0.0002637372],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2327008,"threshold_uncertainty_score":0.9987621,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7400976786095773,"score_gpt":0.5810570523530081,"score_spread":0.1590406262565692,"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."}}