{"id":"W6940649611","doi":"10.1016/j.oceaneng.2025.122163","title":"A pollution occurrence prediction model based on data-driven approaches for Canadian waters","year":2025,"lang":"en","type":"article","venue":"Ocean Engineering","topic":"Mycorrhizal Fungi and Plant Interactions","field":"Agricultural and Biological Sciences","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"","keywords":"Pollution; Missing data; Marine pollution; Imputation (statistics); Random forest; Logistic regression; Air pollution; Bayes' theorem","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00007118381,0.00007522921,0.00005930142,0.00003688259,0.0001113553,0.00003183037,0.0001570808,0.00004498431,0.00002351097],"category_scores_gemma":[0.00005351043,0.00003585007,0.00002815238,0.0001254916,0.000001571245,0.0000907118,0.00001260637,0.00006709732,0.000007250935],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004747446,"about_ca_system_score_gemma":0.00001979915,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00180036,"about_ca_topic_score_gemma":0.001043127,"domain_scores_codex":[0.9995227,0.000005375642,0.00008046329,0.0001793684,0.00004687091,0.0001652333],"domain_scores_gemma":[0.9997046,0.000147265,0.00001188021,0.00006352116,0.00001276708,0.000059944],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002357259,0.00001604888,0.002343202,0.00002228713,0.00001157016,5.303169e-7,0.0000201534,0.9606419,0.002846309,0.001338865,0.02730348,0.005432047],"study_design_scores_gemma":[0.0000237076,0.00002404925,0.0007468668,0.00004805713,0.000009311323,5.245594e-7,0.00001396326,0.9838299,0.0001454335,0.00002394199,0.01506953,0.00006471518],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7363117,0.0001794487,0.1084144,0.008919572,0.003111565,0.002098295,0.1326997,0.0008307125,0.007434585],"genre_scores_gemma":[0.9923941,0.000008018746,0.0004393266,0.0001354231,0.00006107838,0.000008713685,0.006863683,5.280504e-7,0.00008907449],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2560824,"threshold_uncertainty_score":0.2721618,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03153419852844139,"score_gpt":0.2026086824264138,"score_spread":0.1710744838979724,"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."}}