{"id":"W4386999101","doi":"10.1016/j.watres.2023.120667","title":"Machine learning for modeling N2O emissions from wastewater treatment plants: Aligning model performance, complexity, and interpretability","year":2023,"lang":"en","type":"article","venue":"Water Research","topic":"Air Quality Monitoring and Forecasting","field":"Environmental Science","cited_by":96,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université Laval; University of Alberta","funders":"Canada Research Chairs","keywords":"Interpretability; Machine learning; Computer science; Artificial intelligence; Feature selection; Process (computing); Feature (linguistics); Artificial neural network; Model selection; Data mining","routes":{"ca_aff":true,"ca_fund":true,"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":[0.001785246,0.0007191621,0.0005518918,0.0004607517,0.0003895825,0.001070721,0.0005769476,0.0007954335,0.0004007333],"category_scores_gemma":[0.006597665,0.0002584942,0.0006843285,0.0003078082,0.0004194173,0.001118122,0.0005490654,0.0008380022,0.00007262183],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001208018,"about_ca_system_score_gemma":0.0009740559,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01692897,"about_ca_topic_score_gemma":0.01132396,"domain_scores_codex":[0.9994988,0.0002415079,0.00003656563,0.0001097897,0.00006516757,0.00004805045],"domain_scores_gemma":[0.9970208,0.002419682,0.0002275989,0.0001094668,0.0001877962,0.00003465396],"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":[0.00004508368,0.00004871139,0.00330611,0.00001337391,0.00003053719,0.00001455699,0.0000132946,0.9869918,0.000421066,0.0005633752,0.00007708626,0.008474945],"study_design_scores_gemma":[0.000001477182,0.000005338499,0.0002533334,9.033692e-7,0.000001867632,0.000001349306,0.000002139518,0.9990552,0.000131768,0.0005270924,0.00001822389,0.000001216441],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7667381,0.0004042823,0.2292008,0.0007386393,0.00004415149,0.00006389672,0.0002577074,0.000355545,0.002197079],"genre_scores_gemma":[0.9898813,0.00005493484,0.009542113,0.00002281593,0.00001223179,0.00002661131,0.0001304381,0.00001320112,0.000316282],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01692897,"threshold_uncertainty_score":0.03366089,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2353114245475959,"score_gpt":0.377125246212714,"score_spread":0.1418138216651181,"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."}}