{"id":"W2138561153","doi":"10.1029/2006wr005187","title":"Data management of river water quality data: A semi‐automatic procedure for data validation","year":2007,"lang":"en","type":"article","venue":"Water Resources Research","topic":"Fault Detection and Control Systems","field":"Engineering","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université Laval","funders":"","keywords":"Univariate; Outlier; Multivariate statistics; Data mining; Anomaly detection; Data quality; Computer science; Statistics; Mathematics; Artificial intelligence; Engineering","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":[0.01548239,0.00129361,0.001375466,0.005308266,0.001212684,0.002680189,0.002504125,0.001142038,0.002319507],"category_scores_gemma":[0.04229321,0.0008014243,0.001059299,0.002201621,0.001250645,0.002073971,0.00274886,0.002417242,0.00227264],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000600133,"about_ca_system_score_gemma":0.002459806,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001951747,"about_ca_topic_score_gemma":0.001103945,"domain_scores_codex":[0.9857588,0.00575752,0.001748628,0.001928615,0.004462224,0.0003442265],"domain_scores_gemma":[0.952472,0.02232783,0.0037497,0.01297673,0.007974519,0.0004991986],"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":[0.0007056979,0.0006876363,0.01010362,0.0005290377,0.0003800064,0.0003980205,0.001049497,0.02760462,0.06817001,0.01061822,0.01166658,0.8680871],"study_design_scores_gemma":[0.0003513765,0.0007191316,0.01862948,0.0003521239,0.0001534341,0.001060215,0.0003909845,0.7430779,0.1661093,0.024985,0.04365443,0.0005165701],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004903071,0.00004480094,0.9847045,0.0000763201,0.00002517351,0.0003992914,0.0004143851,0.009040697,0.0003918439],"genre_scores_gemma":[0.1111491,0.00005907749,0.8834582,0.0001279033,0.00004580458,0.001765355,0.002124588,0.0006655637,0.0006045366],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01548239,"threshold_uncertainty_score":0.08187973,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1623059939891597,"score_gpt":0.4017803215990002,"score_spread":0.2394743276098405,"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."}}