{"id":"W2147626660","doi":"10.1093/bioinformatics/btr591","title":"Optimized application of penalized regression methods to diverse genomic data","year":2011,"lang":"en","type":"article","venue":"Bioinformatics","topic":"Gene expression and cancer classification","field":"Biochemistry, Genetics and Molecular Biology","cited_by":99,"is_retracted":false,"has_abstract":true,"ca_institutions":"University Health Network; Ontario Institute for Cancer Research","funders":"Division of Biological Infrastructure; National Science Foundation","keywords":"Elastic net regularization; Lasso (programming language); Feature selection; Computer science; Univariate; Regression; Data mining; R package; Predictive modelling; Regression analysis; Machine learning; Artificial intelligence; Statistics; Mathematics; Multivariate statistics","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":[],"consensus_categories":[],"category_scores_codex":[0.005987588,0.0007897026,0.0006406039,0.0006443608,0.0003858375,0.0008337926,0.001663584,0.0009034696,0.00201695],"category_scores_gemma":[0.01962315,0.0003686898,0.0006794261,0.0007909868,0.0007730629,0.0006151317,0.0009985808,0.001610011,0.0006697417],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006573933,"about_ca_system_score_gemma":0.00143908,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003127318,"about_ca_topic_score_gemma":0.003158405,"domain_scores_codex":[0.9972532,0.001763252,0.0001138093,0.0004064724,0.0003583611,0.0001050127],"domain_scores_gemma":[0.9887146,0.008532397,0.0005796468,0.0007568452,0.001252315,0.0001641907],"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":[0.0001361162,0.00006788164,0.002522387,0.0001413435,0.00007066091,0.0001200943,0.00006790586,0.9473239,0.00475012,0.005630082,0.002030392,0.03713918],"study_design_scores_gemma":[0.00001467331,0.00002141931,0.0003380259,0.000009343359,0.000004679217,0.0000215837,0.000006707632,0.994876,0.001624882,0.002559188,0.0005160198,0.000007540169],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02973384,0.000152333,0.9678335,0.0002781015,0.00003119936,0.00006817115,0.0003148551,0.0008971882,0.0006908325],"genre_scores_gemma":[0.2942119,0.0001791083,0.7017221,0.0002727931,0.00004096669,0.0004946465,0.001392765,0.000521627,0.001164147],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005987588,"threshold_uncertainty_score":0.03166574,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08353315231449081,"score_gpt":0.3586795711320192,"score_spread":0.2751464188175284,"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."}}