{"id":"W4382654846","doi":"10.1200/cci.23.00071","title":"Status of Synthetic Data Generation for Structured Health Data","year":2023,"lang":"en","type":"editorial","venue":"JCO Clinical Cancer Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Agricultural Research Institute of Ontario; University of Ottawa","funders":"","keywords":"Synthetic data; Computer science; Data science; Artificial intelligence","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":[],"consensus_categories":[],"category_scores_codex":[0.02424056,0.0008986806,0.00170049,0.002775601,0.001102452,0.006867455,0.003288578,0.00743172,0.007600703],"category_scores_gemma":[0.1159962,0.0007914255,0.00141356,0.001480094,0.003061964,0.004898028,0.002098923,0.01284075,0.004803703],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00170601,"about_ca_system_score_gemma":0.004484975,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001790612,"about_ca_topic_score_gemma":0.003202053,"domain_scores_codex":[0.9902609,0.003617658,0.001218065,0.0008042008,0.003904349,0.0001948605],"domain_scores_gemma":[0.8143514,0.1455894,0.002339827,0.004488461,0.02998701,0.00324396],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008086598,0.00002243301,0.00008566152,0.001168879,0.00008901717,0.000100064,0.00006230955,0.0004442885,0.0001697998,0.007755239,0.8816471,0.1083743],"study_design_scores_gemma":[0.00004078377,0.00002595827,0.00008083014,0.0009548533,0.00005125377,0.0001351392,0.00002139669,0.0009301766,0.0002558233,0.01037441,0.9871034,0.0000259875],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"editorial","genre_gemma":"editorial","genre_scores_codex":[0.0003111981,0.1463778,0.01835895,0.2211838,0.6044733,0.00008362559,0.000686899,0.0008548736,0.00766958],"genre_scores_gemma":[0.006290177,0.1088084,0.01924084,0.09870706,0.7493477,0.0002435204,0.001092179,0.001082414,0.01518766],"genre_candidate":"editorial","genre_consensus":"editorial","teacher_disagreement_score":0.02424056,"threshold_uncertainty_score":0.1281978,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3509988297603718,"score_gpt":0.5450554057956647,"score_spread":0.1940565760352929,"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."}}