{"id":"W3191709143","doi":"10.1093/jamia/ocab128","title":"Improving domain adaptation in de-identification of electronic health records through self-training","year":2021,"lang":"en","type":"article","venue":"Journal of the American Medical Informatics Association","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Google (Canada); University of Toronto","funders":"","keywords":"Computer science; Identification (biology); Software deployment; Domain (mathematical analysis); Domain adaptation; Artificial intelligence; Task (project management); Machine learning; Adaptation (eye); Test data; Data mining; Classifier (UML)","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":[0.003987327,0.001068593,0.001012365,0.001190032,0.0005335639,0.000899347,0.00123551,0.001533656,0.0009847906],"category_scores_gemma":[0.009657885,0.0003047144,0.0009662621,0.0009121977,0.00059847,0.002157612,0.001554061,0.002088989,0.001300422],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006699692,"about_ca_system_score_gemma":0.001118937,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002997626,"about_ca_topic_score_gemma":0.003190667,"domain_scores_codex":[0.9975696,0.00101145,0.0001595346,0.0007803667,0.0003164717,0.000162452],"domain_scores_gemma":[0.9949892,0.002539301,0.0004390905,0.00111889,0.0007775154,0.0001359501],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0006716429,0.001550621,0.02259064,0.0004404973,0.0002499458,0.000466103,0.0004294638,0.1865338,0.02367892,0.003004008,0.02500914,0.7353752],"study_design_scores_gemma":[0.00004812349,0.0001631307,0.004645483,0.00007393908,0.00005349939,0.0003972715,0.0001631712,0.9662744,0.0179277,0.003574262,0.00664198,0.00003703456],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1935665,0.003839857,0.7834167,0.001573794,0.0005915961,0.0003904883,0.001187879,0.01004938,0.005383738],"genre_scores_gemma":[0.7406869,0.001138223,0.2444928,0.001357479,0.0002811172,0.0002576992,0.006268478,0.000239512,0.005277823],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003987327,"threshold_uncertainty_score":0.02108723,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01171142378810457,"score_gpt":0.3038563359614966,"score_spread":0.292144912173392,"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."}}