{"id":"W2172743337","doi":"10.1101/005983","title":"Avoiding test set bias with rank-based prediction","year":2014,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Princess Margaret Cancer Centre; University Health Network; Montreal Clinical Research Institute","funders":"BC Cancer Agency","keywords":"Rank (graph theory); Test (biology); Set (abstract data type); Statistics; Test set; Econometrics; Computer science; Artificial intelligence; Mathematics; Combinatorics; Biology","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":[0.01469431,0.001340253,0.001452034,0.001698459,0.0005062577,0.001695096,0.001654153,0.001003355,0.001432189],"category_scores_gemma":[0.03607006,0.0004139332,0.001204564,0.001292449,0.001331533,0.001327383,0.001457633,0.002096306,0.000790213],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001141848,"about_ca_system_score_gemma":0.001813741,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002258507,"about_ca_topic_score_gemma":0.001960054,"domain_scores_codex":[0.9898669,0.005502005,0.000563372,0.001451971,0.002235841,0.0003799004],"domain_scores_gemma":[0.9651407,0.02349984,0.003244302,0.00460216,0.003089062,0.0004239684],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0007376372,0.0002530518,0.02778946,0.0002179785,0.0005474348,0.0002686882,0.00009341706,0.6865373,0.01266879,0.009576372,0.003950757,0.2573591],"study_design_scores_gemma":[0.00001925324,0.0001582347,0.00188632,0.00001260219,0.00004638777,0.0000616937,0.00001026438,0.979151,0.009367873,0.008600047,0.0006592018,0.00002705414],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05839916,0.0003545076,0.9366137,0.0005774038,0.00006524147,0.0001131538,0.0004440798,0.00236264,0.001070256],"genre_scores_gemma":[0.7774906,0.0001338457,0.2197227,0.0002993592,0.0001364757,0.0001688681,0.0008661316,0.0002480209,0.0009340734],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01469431,"threshold_uncertainty_score":0.07771194,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01906700720823375,"score_gpt":0.2247218676357204,"score_spread":0.2056548604274867,"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."}}