{"id":"W4220808180","doi":"10.1016/j.isprsjprs.2022.02.023","title":"Spatiotemporal distribution of labeled data can bias the validation and selection of supervised learning algorithms: A marine remote sensing example","year":2022,"lang":"en","type":"article","venue":"ISPRS Journal of Photogrammetry and Remote Sensing","topic":"Marine and coastal ecosystems","field":"Earth and Planetary Sciences","cited_by":31,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Algorithm; Computer science; Data set; Remote sensing; Satellite; Matching (statistics); Sampling (signal processing); Set (abstract data type); Data mining; Statistics; Mathematics; Artificial intelligence; Geography","routes":{"ca_aff":true,"ca_fund":false,"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":[0.03069094,0.0007193386,0.0007246587,0.0007958441,0.001087327,0.00124041,0.001411435,0.001467974,0.0005632142],"category_scores_gemma":[0.06809981,0.000340898,0.0009100834,0.0009685046,0.002113535,0.001458531,0.0009012441,0.001273589,0.0002001569],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001176473,"about_ca_system_score_gemma":0.0009355261,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006080536,"about_ca_topic_score_gemma":0.007605819,"domain_scores_codex":[0.9897459,0.006766097,0.000519782,0.001636582,0.001074013,0.0002577179],"domain_scores_gemma":[0.8978534,0.07854956,0.004281544,0.01055188,0.008328887,0.0004348524],"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":[0.001539706,0.0005972275,0.1772087,0.0003110606,0.0007053862,0.0006233408,0.001238196,0.4961853,0.01010025,0.01484946,0.005184386,0.2914569],"study_design_scores_gemma":[0.0001310685,0.0003286744,0.02478543,0.0001011904,0.00008113532,0.0002143889,0.0002942014,0.9358868,0.01256858,0.02316407,0.0023834,0.00006107774],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5190416,0.00122296,0.4732435,0.002094871,0.0001977426,0.0002862839,0.0005262585,0.0005847838,0.002801925],"genre_scores_gemma":[0.9065966,0.000187145,0.09097046,0.0003711997,0.0000551927,0.0001748082,0.0006865152,0.0001175456,0.0008404448],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03069094,"threshold_uncertainty_score":0.1623111,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03792260276340074,"score_gpt":0.2373618406598871,"score_spread":0.1994392378964863,"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."}}