{"id":"W2975926778","doi":"10.1111/gcb.14845","title":"Gap‐filling approaches for eddy covariance methane fluxes: A comparison of three machine learning algorithms and a traditional method with principal component analysis","year":2019,"lang":"en","type":"article","venue":"Global Change Biology","topic":"Atmospheric and Environmental Gas Dynamics","field":"Environmental Science","cited_by":209,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"British Columbia Knowledge Development Fund; Natural Sciences and Engineering Research Council of Canada; Ministry of Science and ICT, South Korea; Canadian Network for Research and Innovation in Machining Technology, Natural Sciences and Engineering Research Council of Canada; Department of Water Resources; National Research Foundation of Korea; Canada Foundation for Innovation","keywords":"Eddy covariance; Principal component analysis; Algorithm; Methane; Covariance; Component (thermodynamics); Computer science; Component analysis; Environmental science; Pattern recognition (psychology); Artificial intelligence; Mathematics; Chemistry; Ecosystem; Statistics; Physics","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":[0.0002658794,0.0001866364,0.0004803585,0.00001147071,0.00006580599,0.000006155682,0.0001465799,0.0001116143,0.0001622081],"category_scores_gemma":[0.000007302427,0.0001484223,0.0000868442,0.0002628102,0.0002412639,0.00006714817,0.0001253128,0.0001070446,0.000005915868],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001370206,"about_ca_system_score_gemma":0.000004512444,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00106118,"about_ca_topic_score_gemma":0.0002650513,"domain_scores_codex":[0.9988147,0.00008684973,0.0002322352,0.0004360671,0.0001517218,0.0002783973],"domain_scores_gemma":[0.9994851,0.00009139112,0.0001975071,0.0001392225,0.000004391164,0.00008239398],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001722854,0.0001432892,0.910632,0.00002159774,0.0002930861,7.143705e-7,0.0002663485,0.07080054,0.0002103989,0.000845751,0.000001412803,0.01661261],"study_design_scores_gemma":[0.0006737886,0.0006610176,0.1789868,0.000006094295,0.0002312942,0.00001122118,0.0002029062,0.8179967,0.00002047679,0.0004687381,0.0005522483,0.000188662],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4290474,0.0002360529,0.5700461,0.00006158475,0.00003623391,0.0003570734,0.00008623586,0.00001446551,0.0001148033],"genre_scores_gemma":[0.6798264,0.00001633028,0.3198521,0.00003983331,0.00002437325,0.00005055125,0.0001612741,0.000006809513,0.0000223411],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7471962,"threshold_uncertainty_score":0.6052485,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1005651676502593,"score_gpt":0.2903182678598666,"score_spread":0.1897531002096073,"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."}}