{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.2615435,0.002238546,0.002301304,0.006180553,0.001895794,0.006104504,0.003839041,0.0021196,0.005492392],"category_scores_gemma":[0.3659699,0.001439797,0.003918918,0.006553713,0.002138332,0.003037203,0.009449063,0.003773621,0.00118236],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002499841,"about_ca_system_score_gemma":0.007689696,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006974491,"about_ca_topic_score_gemma":0.008782038,"domain_scores_codex":[0.8622901,0.1081801,0.01019254,0.006101145,0.01236273,0.0008734133],"domain_scores_gemma":[0.5041586,0.3964302,0.02173614,0.04084805,0.03268385,0.004143118],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001723764,0.0006197796,0.07743312,0.00304588,0.003286277,0.000982009,0.007020927,0.0467605,0.005922165,0.03100567,0.03319997,0.789],"study_design_scores_gemma":[0.001790312,0.001849231,0.05663292,0.001936836,0.001231013,0.001523057,0.002435556,0.7527584,0.01267073,0.09719331,0.06943769,0.0005409335],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01273922,0.0006076916,0.9723809,0.002390414,0.0001413863,0.001933636,0.001315704,0.007149385,0.001341747],"genre_scores_gemma":[0.02710107,0.0001528849,0.9698727,0.0001556511,0.00003819018,0.001258569,0.000458668,0.0007107194,0.0002515957],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7384565,"threshold_uncertainty_score":0.9106484,"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."}}