{"id":"W2946571535","doi":"10.1186/s12874-019-0737-5","title":"The utility of multivariate outlier detection techniques for data quality evaluation in large studies: an application within the ONDRI project","year":2019,"lang":"en","type":"article","venue":"BMC Medical Research Methodology","topic":"Dementia and Cognitive Impairment Research","field":"Medicine","cited_by":72,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Ontario; Canada Research Chairs; Health Sciences Centre; University of New Brunswick; University of Toronto; Western University; Sunnybrook Health Science Centre; Lawson Health Research Institute; Parkwood Institute; University of Waterloo; McMaster University; Baycrest Hospital","funders":"Ontario Ministry of Research and Innovation; Western University; University of Toronto; Strong; Canadian Institutes of Health Research; Sunnybrook Research Institute; Ontario Brain Institute; Ontario Ministry of Research, Innovation and Science; Heart and Stroke Foundation of Canada","keywords":"Multivariate statistics; Computer science; Outlier; Multivariate analysis; Data mining; Anomaly detection; Data science; Quality (philosophy); Data quality; Information retrieval; Machine learning; Artificial intelligence; Operations management; Engineering","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.2782478,0.0001146248,0.0004308729,0.0002120991,0.0002330791,0.00001718079,0.0007208892,0.0002093983,0.00008891346],"category_scores_gemma":[0.1859597,0.00005966346,0.00006245358,0.0006119853,0.000699095,0.0001211889,0.0005684907,0.0008338876,0.000006885384],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001257888,"about_ca_system_score_gemma":0.001922663,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001072072,"about_ca_topic_score_gemma":0.008421349,"domain_scores_codex":[0.9571357,0.03709046,0.0009617342,0.0008302447,0.003274126,0.0007077698],"domain_scores_gemma":[0.9411774,0.05377441,0.0002711334,0.001961683,0.002637554,0.0001778724],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.007655292,0.001122362,0.1054026,0.001101965,0.0002065182,0.000001613068,0.002231303,3.833231e-7,0.02207871,0.003614688,0.0003087677,0.8562758],"study_design_scores_gemma":[0.007630313,0.003630701,0.4589725,0.0002930778,0.0001408462,0.00001445386,0.02413337,0.4392225,0.03490026,0.02145493,0.009399719,0.000207309],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.647819,0.000496122,0.337465,0.001819121,0.0001644867,0.01166635,0.00002601114,0.00003515748,0.0005087331],"genre_scores_gemma":[0.9804097,0.0001948864,0.01635857,0.0001038593,0.000121554,0.002469776,0.0001122118,0.00001513267,0.0002142654],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8560685,"threshold_uncertainty_score":0.8208973,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7549132218594857,"score_gpt":0.6851126797262486,"score_spread":0.06980054213323716,"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."}}