{"id":"W1602160603","doi":"10.1186/1471-2288-14-137","title":"Modern modelling techniques are data hungry: a simulation study for predicting dichotomous endpoints","year":2014,"lang":"en","type":"article","venue":"BMC Medical Research Methodology","topic":"Traumatic Brain Injury and Neurovascular Disturbances","field":"Medicine","cited_by":796,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Support vector machine; Logistic regression; Statistics; Random forest; Sample size determination; Artificial intelligence; Receiver operating characteristic; Regression; Machine learning; Computer science; Medicine; Mathematics","routes":{"ca_aff":true,"ca_fund":false,"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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01343246,0.0009132687,0.0008985709,0.0006925067,0.0003526072,0.0007144157,0.001079654,0.001457126,0.0008015174],"category_scores_gemma":[0.03000661,0.0004963013,0.001895121,0.0006503568,0.0008278792,0.0008297442,0.0008087996,0.001519254,0.00009440615],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001217451,"about_ca_system_score_gemma":0.0009293354,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005871278,"about_ca_topic_score_gemma":0.003358271,"domain_scores_codex":[0.9952348,0.003996449,0.0001519956,0.0002608678,0.0001959876,0.0001598836],"domain_scores_gemma":[0.9150562,0.07786722,0.002645739,0.002317831,0.001518498,0.0005945324],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001602138,0.0005907543,0.05221243,0.0001627764,0.0006143012,0.0001864097,0.0001826096,0.9349462,0.0006115954,0.001611839,0.0003218239,0.006957042],"study_design_scores_gemma":[0.0002145076,0.00116876,0.007404317,0.00004127735,0.000145879,0.00009352786,0.0000649696,0.9884921,0.0006333607,0.001521397,0.0001894269,0.00003046252],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9781291,0.0004023379,0.02028446,0.000254532,0.00001919017,0.0001345488,0.000272587,0.00004114769,0.0004621391],"genre_scores_gemma":[0.9889007,0.000147637,0.01027618,0.00005371751,0.00001091621,0.0001931677,0.0002567469,0.000006538243,0.0001543914],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9865676,"threshold_uncertainty_score":0.07103848,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7132462483034406,"score_gpt":0.5762959983442477,"score_spread":0.1369502499591929,"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."}}