{"id":"W4379058118","doi":"10.1016/j.geoderma.2023.116553","title":"Divergence metrics for determining optimal training sample size in digital soil mapping","year":2023,"lang":"en","type":"article","venue":"Geoderma","topic":"Soil Geostatistics and Mapping","field":"Environmental Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph; Ministry of Agriculture, Food and Rural Affairs","funders":"Natural Sciences and Engineering Research Council of Canada; Ontario Ministry of Agriculture, Food and Rural Affairs","keywords":"Sample size determination; Statistics; Covariate; Mathematics; Divergence (linguistics); Sample (material); Sampling (signal processing); Computer science","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":[0.03352533,0.001147891,0.001905822,0.002703208,0.001126295,0.001617578,0.002136009,0.002248831,0.00107621],"category_scores_gemma":[0.1242607,0.0006524042,0.0009589038,0.001854167,0.002005237,0.002445661,0.002539042,0.002710047,0.0002545374],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001788568,"about_ca_system_score_gemma":0.002438194,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003804274,"about_ca_topic_score_gemma":0.004323857,"domain_scores_codex":[0.988315,0.006519607,0.001369837,0.001253868,0.002219409,0.0003222924],"domain_scores_gemma":[0.9067155,0.08031121,0.003064418,0.002736417,0.006527196,0.000645273],"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":[0.000828254,0.0004652946,0.03735767,0.0007682169,0.0004033521,0.0003654319,0.000667168,0.6451856,0.01102475,0.01952392,0.002477705,0.2809327],"study_design_scores_gemma":[0.00009654698,0.0004224776,0.008399691,0.0001458229,0.00004715256,0.0001934655,0.0001953737,0.9652169,0.00914204,0.01471015,0.001347905,0.0000824242],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.09644492,0.000822773,0.8995308,0.0003054319,0.00005450572,0.0003859933,0.0003390026,0.0009179843,0.001198688],"genre_scores_gemma":[0.3172839,0.0001984102,0.6803567,0.0001549315,0.0000276698,0.0007873465,0.0006760448,0.0002153033,0.000299726],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.03352533,"threshold_uncertainty_score":0.177301,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04843698902770568,"score_gpt":0.2588117446673414,"score_spread":0.2103747556396357,"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."}}