{"id":"W4304208350","doi":"10.1093/jamiaopen/ooac083","title":"Validating a membership disclosure metric for synthetic health data","year":2022,"lang":"en","type":"article","venue":"JAMIA Open","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":38,"is_retracted":false,"has_abstract":true,"ca_institutions":"Children's Hospital of Eastern Ontario; University of Ottawa","funders":"Canadian Institutes of Health Research","keywords":"Synthetic data; Metric (unit); Computer science; Benchmark (surveying); Population; Data mining; Generative model; Measure (data warehouse); Fraction (chemistry); Parametrization (atmospheric modeling); Sampling (signal processing); Ground truth; Machine learning; Generative grammar; Artificial intelligence; Geography; Medicine","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.0287102,0.0007835194,0.0005969159,0.001214748,0.0008189225,0.001671072,0.00164592,0.001785685,0.000993193],"category_scores_gemma":[0.1063014,0.0002608011,0.0007124124,0.0008480089,0.00293181,0.002349365,0.002885741,0.001933377,0.0001825195],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00272515,"about_ca_system_score_gemma":0.001627637,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001372951,"about_ca_topic_score_gemma":0.0008289851,"domain_scores_codex":[0.982942,0.0110569,0.000745902,0.001642026,0.00318403,0.0004291213],"domain_scores_gemma":[0.8821571,0.08916667,0.007569976,0.01619253,0.003933146,0.0009807032],"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":[0.0007526518,0.0002540756,0.03268949,0.0002730667,0.0001925869,0.0001532157,0.0005190556,0.877496,0.004621266,0.02955252,0.002354435,0.05114171],"study_design_scores_gemma":[0.00004402576,0.0004174038,0.004615332,0.00007216944,0.00002701053,0.0002737785,0.0001194144,0.9583868,0.009434548,0.02527406,0.00130042,0.00003502843],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5142393,0.0007648906,0.4763696,0.002271696,0.0001262185,0.0005598853,0.001392608,0.0006512236,0.003624671],"genre_scores_gemma":[0.9573352,0.00008521627,0.04136573,0.0001665018,0.00002234262,0.0001352561,0.0006363096,0.0000239752,0.000229375],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0287102,"threshold_uncertainty_score":0.1518359,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1772611440591591,"score_gpt":0.3833628589269482,"score_spread":0.2061017148677891,"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."}}