{"id":"W4412195443","doi":"10.1002/env.70023","title":"Occupancy Modeling for Rare Species Using Large Datasets: A Subsampling Approach","year":2025,"lang":"en","type":"article","venue":"Environmetrics","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Actua; Western University","funders":"Natural Sciences and Engineering Research Council of Canada; Institute for Catastrophic Loss Reduction; Canadian Statistical Sciences Institute; Ontario Ministry of Natural Resources and Forestry","keywords":"Occupancy; Inference; Covariate; Negative binomial distribution; Computer science; Statistics; Sampling (signal processing); Distance sampling; Abundance (ecology); Ecology; Artificial intelligence; Machine learning; Mathematics; Poisson distribution; Biology","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.01655864,0.0009007998,0.001435984,0.001805614,0.001277285,0.001543119,0.003084343,0.001349206,0.001346181],"category_scores_gemma":[0.03155232,0.0009552913,0.00205116,0.001636082,0.001106458,0.001523086,0.001692611,0.001849536,0.0002366176],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001501795,"about_ca_system_score_gemma":0.001290323,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04967454,"about_ca_topic_score_gemma":0.05979912,"domain_scores_codex":[0.9941576,0.004207787,0.0002289686,0.0009189855,0.0003095918,0.0001771034],"domain_scores_gemma":[0.9699435,0.02423332,0.001647269,0.002836028,0.0009593607,0.0003805089],"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":[0.0005387932,0.0004987391,0.1625791,0.0002303248,0.001212315,0.0005012035,0.0007076429,0.7414109,0.001478745,0.01570374,0.004579504,0.070559],"study_design_scores_gemma":[0.00002816905,0.00003997593,0.006428612,0.00001739129,0.00004467666,0.00005344587,0.00008558972,0.9844669,0.0001788806,0.007730146,0.0009073005,0.00001904592],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2146973,0.0006188487,0.7801027,0.0008995256,0.0001011461,0.0003112782,0.001856921,0.0007027179,0.0007095666],"genre_scores_gemma":[0.7597561,0.0002549826,0.2338657,0.0003570149,0.0001892688,0.0005601929,0.00385777,0.0001124867,0.001046496],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04967454,"threshold_uncertainty_score":0.0987708,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08522441394924438,"score_gpt":0.2984696268860576,"score_spread":0.2132452129368132,"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."}}