{"id":"W4401373758","doi":"10.1214/24-aoas1877","title":"Probabilistic contrastive dimension reduction for case-control study data","year":2024,"lang":"en","type":"article","venue":"The Annals of Applied Statistics","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Center for Advancing Translational Sciences; National Human Genome Research Institute; National Heart, Lung, and Blood Institute; Canadian Institute for Advanced Research; National Cancer Institute; National Science Foundation; National Institutes of Health; National Institute of Environmental Health Sciences; Leona M. and Harry B. Helmsley Charitable Trust","keywords":"Dimensionality reduction; Dimension (graph theory); Reduction (mathematics); Probabilistic logic; Computer science; Statistics; Data reduction; Artificial intelligence; Natural language processing; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04075399,0.001089864,0.001597415,0.002722795,0.001045747,0.001870676,0.002595002,0.001959879,0.001448242],"category_scores_gemma":[0.1160982,0.0009046886,0.002394829,0.002099454,0.00382221,0.001496007,0.002352076,0.003964915,0.0003355113],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001673081,"about_ca_system_score_gemma":0.001877899,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002266629,"about_ca_topic_score_gemma":0.001822624,"domain_scores_codex":[0.9719257,0.02164065,0.0009270888,0.002969757,0.002326252,0.0002105169],"domain_scores_gemma":[0.9125114,0.0731402,0.003472788,0.008529861,0.001984606,0.0003611689],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001119371,0.0003945045,0.02456483,0.001048868,0.001717549,0.0008310105,0.001001684,0.3031078,0.009804653,0.2836433,0.01037278,0.3623936],"study_design_scores_gemma":[0.0001322687,0.0001561858,0.005128036,0.00007972323,0.0001196484,0.0003577184,0.00004702397,0.7510155,0.00262396,0.2353159,0.004952166,0.00007181002],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005045044,0.0003822597,0.9934002,0.0004080937,0.00003466155,0.0001743495,0.0001587341,0.0002083337,0.0001883516],"genre_scores_gemma":[0.2339985,0.0005284492,0.76179,0.0005510615,0.0002187273,0.001435867,0.0009172196,0.0001094668,0.000450689],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04075399,"threshold_uncertainty_score":0.2155303,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08856121255749513,"score_gpt":0.3719493295497239,"score_spread":0.2833881169922288,"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."}}