{"id":"W3033682799","doi":"10.2196/18910","title":"Reliability of Supervised Machine Learning Using Synthetic Data in Health Care: Model to Preserve Privacy for Data Sharing","year":2020,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":164,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Horizon 2020 Framework Programme; European Commission","keywords":"Synthetic data; Machine learning; Computer science; Artificial intelligence; Random forest; Decision tree; Support vector machine; Generalizability theory; Data mining; Statistics; Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02823386,0.0006905115,0.0009824767,0.0009083957,0.0005710328,0.001532384,0.001778458,0.001354033,0.0008444533],"category_scores_gemma":[0.08468535,0.000401442,0.0008568791,0.0007747887,0.001934636,0.002430261,0.00148459,0.001981408,0.0002582919],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002168765,"about_ca_system_score_gemma":0.001610875,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004339682,"about_ca_topic_score_gemma":0.002476961,"domain_scores_codex":[0.986833,0.0094915,0.0005453914,0.001429745,0.001329308,0.0003710354],"domain_scores_gemma":[0.9002564,0.07377248,0.007463906,0.01025585,0.007538681,0.000712678],"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":[0.000600511,0.0002285244,0.02303099,0.0001359263,0.0001386014,0.0000938068,0.0002801369,0.9254795,0.000847109,0.007776391,0.001347109,0.04004135],"study_design_scores_gemma":[0.00001197932,0.00008421523,0.001392686,0.00001743194,0.000009631389,0.0000332317,0.00003051476,0.9921203,0.0007086743,0.005362709,0.0002197737,0.000008955653],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.5552764,0.000816383,0.4364639,0.002711921,0.0001419501,0.0004223186,0.0007307637,0.0006011262,0.002835226],"genre_scores_gemma":[0.9716732,0.00009203424,0.02713172,0.000132436,0.00003173985,0.000128224,0.0004248637,0.00002221808,0.0003635988],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.02823386,"threshold_uncertainty_score":0.1493167,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.152395735813922,"score_gpt":0.379612199091731,"score_spread":0.227216463277809,"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."}}