{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01020144,0.001267822,0.001062594,0.001200387,0.0007739877,0.001150054,0.002066784,0.00149397,0.003276072],"category_scores_gemma":[0.03036317,0.0005504996,0.0007231791,0.001057705,0.0006582826,0.002128684,0.0009425702,0.002622959,0.001836616],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004978004,"about_ca_system_score_gemma":0.001741394,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007379159,"about_ca_topic_score_gemma":0.01077691,"domain_scores_codex":[0.9976502,0.001538783,0.0001355044,0.0002894995,0.0002260621,0.0001599516],"domain_scores_gemma":[0.9920015,0.005681112,0.000381434,0.0007992897,0.0009355879,0.0002010134],"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":[0.0005694492,0.000446308,0.01617175,0.0003508094,0.0002748896,0.0001418762,0.0001676334,0.5545938,0.005548008,0.004680117,0.0217762,0.3952791],"study_design_scores_gemma":[0.00009107096,0.00005770468,0.001794972,0.0001074149,0.00004075083,0.00003737559,0.00007041158,0.9859149,0.003354782,0.005735952,0.002770726,0.00002382999],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1312036,0.003047321,0.8491649,0.001821712,0.0003346088,0.0004294218,0.000841989,0.008965013,0.004191425],"genre_scores_gemma":[0.5427577,0.0005932684,0.4512473,0.0006765294,0.00014534,0.0004316545,0.00154795,0.001305862,0.001294371],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01020144,"threshold_uncertainty_score":0.05395103,"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."}}