{"id":"W2952088026","doi":"10.2196/14499","title":"Projection Word Embedding Model With Hybrid Sampling Training for Classifying ICD-10-CM Codes: Longitudinal Observational Study","year":2019,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Word2vec; Word embedding; Computer science; Artificial intelligence; Natural language processing; Diagnosis code; Vocabulary; Word (group theory); Data mining; Embedding; Information retrieval; Machine learning; Medicine; Mathematics; Linguistics","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":[0.008116547,0.0009454911,0.0007976901,0.0007884413,0.0005697529,0.0006180978,0.0009169923,0.0006346412,0.002139251],"category_scores_gemma":[0.01401289,0.0004241598,0.001396991,0.0007857857,0.0004045474,0.001231478,0.0008032098,0.001750352,0.0003876289],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005425306,"about_ca_system_score_gemma":0.001012218,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0116368,"about_ca_topic_score_gemma":0.009484924,"domain_scores_codex":[0.9980106,0.001225643,0.0001268937,0.0004000369,0.0001238106,0.00011303],"domain_scores_gemma":[0.9928088,0.00408731,0.000736238,0.0008233514,0.001210092,0.000334116],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.002508617,0.002087994,0.8812473,0.0003467973,0.001401155,0.000280489,0.0007206704,0.02539756,0.0007961764,0.001041302,0.006723563,0.07744838],"study_design_scores_gemma":[0.0003328057,0.002989149,0.1712597,0.0002017842,0.001691047,0.0003259448,0.001587006,0.8142517,0.001448266,0.003420281,0.002371999,0.000120336],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9569542,0.0008115126,0.03966058,0.0003193216,0.0001389034,0.0002603065,0.001490459,0.00007581877,0.0002888415],"genre_scores_gemma":[0.9837271,0.0003633362,0.01174422,0.00008728146,0.00007018635,0.0003550109,0.003195885,0.0000191337,0.0004377082],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0116368,"threshold_uncertainty_score":0.04292488,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1962135696598163,"score_gpt":0.4181820182667698,"score_spread":0.2219684486069535,"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."}}