{"id":"W7123923582","doi":"10.2166/nh.2025.075","title":"How to out-perform default random forest regression: choosing hyperparameters for applications in large-sample hydrology","year":2025,"lang":"en","type":"article","venue":"Hydrology research","topic":"Hydrological Forecasting Using AI","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Regent College; University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Random forest; Leverage (statistics); Hyperparameter; Water resources; Hydrological modelling; Deep learning","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003090192,0.0002526995,0.0004815192,0.0004550834,0.0007494714,0.00008654402,0.0008605559,0.0004123522,0.0002054961],"category_scores_gemma":[0.003547798,0.0002067871,0.0001168026,0.001206987,0.0006917198,0.0001391098,0.0008045089,0.0008011904,0.0002031344],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000352884,"about_ca_system_score_gemma":0.0000533721,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003509937,"about_ca_topic_score_gemma":0.002745605,"domain_scores_codex":[0.9962243,0.0004392958,0.0003884941,0.0009993336,0.0004012977,0.001547277],"domain_scores_gemma":[0.9958737,0.003070798,0.00006591286,0.0006970655,0.00005171189,0.0002408066],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.002141193,0.0008150188,0.8740936,0.00008507171,0.00007019508,0.00003290583,0.001421171,0.07473069,0.01280102,0.001749047,0.01182891,0.02023119],"study_design_scores_gemma":[0.008389751,0.001608398,0.03720668,0.0001154911,0.00005417632,0.00002146116,0.0002259622,0.3957259,0.002923785,0.0683917,0.4844671,0.0008694949],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9541869,0.0000682326,0.02784134,0.0128945,0.0001683832,0.002062377,0.0000248656,0.00008528778,0.002668125],"genre_scores_gemma":[0.9898751,0.000009474453,0.005509067,0.001179051,0.00006806907,0.001657441,0.00004932128,0.00002783012,0.001624632],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8368869,"threshold_uncertainty_score":0.8432535,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06728013377302482,"score_gpt":0.3755970337498849,"score_spread":0.3083168999768601,"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."}}