{"id":"W4410062502","doi":"10.1101/2025.05.02.25326890","title":"A Quantitative Comparison of Structural and Distributional Properties of Synthetic Tabular Data in Parkinson’s Disease","year":2025,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Bayesian Methods and Mixture Models","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"Canadian Institutes of Health Research","keywords":"Parkinson's disease; Disease; Econometrics; Psychology; Computer science; Economics; Medicine; Pathology","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":[],"consensus_categories":[],"category_scores_codex":[0.0005836107,0.0001857314,0.0005341378,0.0001289201,0.00002961108,0.00003420405,0.00137114,0.00008776489,0.000002047826],"category_scores_gemma":[0.000347761,0.0001490258,0.00005074146,0.0001643814,0.0002055153,0.000128618,0.002533251,0.0002797299,1.507064e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002174253,"about_ca_system_score_gemma":0.0002909499,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001045082,"about_ca_topic_score_gemma":0.00002901748,"domain_scores_codex":[0.998141,0.0003312782,0.0004835473,0.0006149563,0.000275416,0.0001538209],"domain_scores_gemma":[0.9982712,0.0001249394,0.000263713,0.001166667,0.0001028512,0.00007068783],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003283807,0.0003721164,0.483112,0.00486854,0.0002119549,0.00002493972,0.004721615,0.0006816154,0.002380274,0.4792826,0.00008382325,0.02393213],"study_design_scores_gemma":[0.0002711168,0.00004371388,0.1674646,0.001559592,0.00006429329,9.720961e-7,0.00004483509,0.7686274,0.003718613,0.05777181,0.0001371611,0.0002959186],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4018118,0.004769108,0.592337,0.0003092916,0.0001504216,0.0002012407,0.0003977556,0.000010513,0.00001283657],"genre_scores_gemma":[0.8772413,0.00004830951,0.1226226,0.000008468884,0.000008428461,0.00001479448,0.00004294899,0.00000350168,0.000009707044],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7679458,"threshold_uncertainty_score":0.6077094,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07936323153758752,"score_gpt":0.3460118988550197,"score_spread":0.2666486673174321,"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."}}