{"id":"W4220991108","doi":"10.1214/22-ejs1998","title":"Adaptive threshold-based classification of sparse high-dimensional data","year":2022,"lang":"en","type":"article","venue":"Electronic Journal of Statistics","topic":"Random Matrices and Applications","field":"Mathematics","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Mathematics; Classifier (UML); Pattern recognition (psychology); Sample size determination; Binary classification; Feature (linguistics); Binary number; Diagonal; Covariance matrix; Linear classifier; Algorithm; Covariance; Artificial intelligence; Statistics; 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":[0.002799846,0.0004606647,0.001487709,0.0009344087,0.0003396551,0.001187965,0.001593616,0.00144026,0.0008680943],"category_scores_gemma":[0.01391477,0.0003398039,0.0005569882,0.001199178,0.001181788,0.001937769,0.001205981,0.001313477,0.0004914796],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006856456,"about_ca_system_score_gemma":0.0007417267,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001274868,"about_ca_topic_score_gemma":0.001017706,"domain_scores_codex":[0.9988652,0.0003533677,0.00008197389,0.0002194784,0.0003754157,0.0001045107],"domain_scores_gemma":[0.9937732,0.00425458,0.0006161994,0.0005932596,0.0006214032,0.000141539],"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":[0.000377353,0.0001704569,0.003504939,0.0002435732,0.0001012306,0.0001780349,0.0002080539,0.7235386,0.02666644,0.02985958,0.001572949,0.2135788],"study_design_scores_gemma":[0.000004603753,0.00002439046,0.0002456759,0.000002982976,0.000003624775,0.00001949769,0.000006492382,0.9937288,0.00140876,0.004415396,0.0001350007,0.000004843201],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02421539,0.0001142623,0.9751604,0.0001223226,0.00001586313,0.00002256103,0.00003236585,0.0001244642,0.0001923583],"genre_scores_gemma":[0.591283,0.0003943215,0.4058369,0.0001792113,0.0001405774,0.0001893437,0.00034107,0.00007497361,0.001560631],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002799846,"threshold_uncertainty_score":0.01480716,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07747682503576564,"score_gpt":0.3228351643396283,"score_spread":0.2453583393038627,"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."}}