{"id":"W3206851169","doi":"10.1145/3469035","title":"Differentially Private Medical Texts Generation Using Generative Neural Networks","year":2021,"lang":"en","type":"article","venue":"ACM Transactions on Computing for Healthcare","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Regina; University of Manitoba","funders":"National Institutes of Health; University of Texas Health Science Center at Houston","keywords":"Computer science; Generative grammar; Medical record; Health care; Data science; Information retrieval; Health records; Big data; Private information retrieval; Patient care; Volume (thermodynamics); Artificial intelligence; Data mining; Medicine; Nursing; Computer security","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","sts"],"consensus_categories":[],"category_scores_codex":[0.0005348461,0.0003078271,0.0003772744,0.0001500053,0.001421412,0.0002317091,0.0009259406,0.0002941012,0.00002859449],"category_scores_gemma":[0.0002694561,0.0003210196,0.0002034289,0.0006150647,0.00004486027,0.0002079839,0.00007951158,0.0009467897,0.000003306324],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002302932,"about_ca_system_score_gemma":0.000530261,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001935226,"about_ca_topic_score_gemma":0.000513699,"domain_scores_codex":[0.996376,0.0006896942,0.0006617521,0.000924482,0.0006409613,0.0007071061],"domain_scores_gemma":[0.9972392,0.0005850157,0.0002077276,0.001078818,0.000488131,0.0004010477],"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":[0.00001927506,0.0001370326,0.001054776,0.000188596,0.00005265792,0.00004489895,0.0007644173,0.4774208,0.0001672374,0.007137239,0.00006142644,0.5129517],"study_design_scores_gemma":[0.0005108842,0.0001985367,0.00086111,0.0001170505,0.00001275729,0.0001249872,0.0000173853,0.9962171,0.0006614055,0.0007238446,0.0002767408,0.0002782505],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1005066,0.000358455,0.8737972,0.02236632,0.002198477,0.0004417059,0.0000093709,0.0003174375,0.000004476721],"genre_scores_gemma":[0.7977951,0.00002834429,0.198321,0.002981042,0.0007308577,0.00002324637,0.00005275567,0.00003730809,0.00003037792],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6972886,"threshold_uncertainty_score":0.9999242,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06312692867739636,"score_gpt":0.3529737858752012,"score_spread":0.2898468571978049,"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."}}