{"id":"W2066942162","doi":"10.1111/j.1751-908x.2010.00084.x","title":"How Fit Are Your Data?","year":2010,"lang":"en","type":"article","venue":"Geostandards and Geoanalytical Research","topic":"Geochemistry and Geologic Mapping","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Chicoutimi","funders":"","keywords":"USable; Quality (philosophy); Data quality; Computer science; Context (archaeology); Sampling (signal processing); Set (abstract data type); Field (mathematics); Data mining; Data set; Risk analysis (engineering); Data science; Operations research; Mathematics; Geology; Artificial intelligence; Operations management; 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":[0.1327069,0.002147695,0.004134765,0.01123808,0.002292777,0.02207321,0.00582882,0.01069748,0.04627591],"category_scores_gemma":[0.5651293,0.002132781,0.003304757,0.01220962,0.004841167,0.03470562,0.005955062,0.01100265,0.06708571],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003522165,"about_ca_system_score_gemma":0.007523576,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003155351,"about_ca_topic_score_gemma":0.002970768,"domain_scores_codex":[0.837285,0.09041914,0.02179585,0.01091821,0.03766302,0.001918687],"domain_scores_gemma":[0.5725819,0.2094584,0.02701637,0.0975517,0.08368126,0.009710387],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0007013775,0.0002942563,0.01194327,0.004836109,0.001068151,0.0004934845,0.001782328,0.0009723386,0.001000838,0.03706791,0.5994011,0.3404388],"study_design_scores_gemma":[0.000166764,0.0001653508,0.003524181,0.006844505,0.0003333106,0.0006892075,0.002724858,0.00166568,0.001585351,0.07724227,0.9047492,0.0003093281],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"commentary","genre_gemma":"methods","genre_scores_codex":[0.008357617,0.02365042,0.19493,0.6147119,0.03434889,0.001981429,0.04895798,0.01325175,0.05980994],"genre_scores_gemma":[0.137485,0.03938841,0.5273874,0.1771484,0.01781508,0.006625662,0.05259595,0.01347755,0.02807653],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.1327069,"threshold_uncertainty_score":0.7018294,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1717799317968001,"score_gpt":0.4028962988718554,"score_spread":0.2311163670750553,"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."}}