{"id":"W4224311959","doi":"10.2196/35734","title":"Utility Metrics for Evaluating Synthetic Health Data Generation Methods: Validation Study","year":2022,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":68,"is_retracted":false,"has_abstract":true,"ca_institutions":"Children's Hospital of Eastern Ontario; University of Ottawa","funders":"","keywords":"Computer science; Metric (unit); Data mining; Machine learning; Synthetic data; Artificial intelligence; Data set","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"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":[0.02370087,0.0001504624,0.0003131044,0.0002405722,0.0008868961,0.0001568602,0.002426493,0.00006162605,0.0001271135],"category_scores_gemma":[0.005772573,0.0001416606,0.00004476284,0.001019881,0.00002755875,0.0006797909,0.00221774,0.0006719258,0.000007095623],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001903566,"about_ca_system_score_gemma":0.001068453,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008034961,"about_ca_topic_score_gemma":0.00001230293,"domain_scores_codex":[0.9933541,0.00222289,0.001373516,0.0003701785,0.002267388,0.0004118927],"domain_scores_gemma":[0.9957455,0.001348511,0.0006580424,0.00177744,0.0001874573,0.0002829997],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000006519856,0.000291532,0.001161175,0.000285424,0.00002075423,8.147243e-7,0.01502098,0.0009693003,0.000001371623,0.001930791,0.004142417,0.9761689],"study_design_scores_gemma":[0.00045851,0.0007197545,0.0003887166,0.00001117624,0.000008256378,0.00001408138,0.002090553,0.9869586,0.000007603039,0.0002634914,0.008952445,0.0001268662],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02161196,0.00008544786,0.9716575,0.003323493,0.0008430562,0.002146339,0.00003664684,0.0002083771,0.00008718017],"genre_scores_gemma":[0.2800863,0.000006299525,0.7162942,0.002466916,0.0001386332,0.0005360399,0.0004290354,0.00001462132,0.0000279485],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9859893,"threshold_uncertainty_score":0.8214297,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3011521977899394,"score_gpt":0.5364280745162925,"score_spread":0.2352758767263531,"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."}}