{"id":"W4286009680","doi":"10.1101/2022.07.18.500262","title":"Class imbalance should not throw you off balance: Choosing the right classifiers and performance metrics for brain decoding with imbalanced data","year":2022,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Imbalanced Data Classification Techniques","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute; McGill University; Concordia University; University of Alberta; Université de Montréal","funders":"Fonds de recherche du Québec – Nature et technologies; Courtois Foundation; Natural Sciences and Engineering Research Council of Canada; Institut de Valorisation des Données; Canada Research Chairs; Canada First Research Excellence Fund; Canadian Institutes of Health Research; Mitacs; McGill University","keywords":"Computer science; Magnetoencephalography; Metric (unit); Artificial intelligence; Robustness (evolution); Binary classification; Machine learning; Decoding methods; Random forest; Class (philosophy); Receiver operating characteristic; Performance metric; Electroencephalography; Binary number; Sensitivity (control systems); Pattern recognition (psychology); Support vector machine; Mathematics; Psychology; Algorithm","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.003448793,0.000955235,0.0009615654,0.0005138009,0.001236439,0.001370261,0.007506569,0.0004946852,0.0000146438],"category_scores_gemma":[0.0007489779,0.0007964838,0.0001193057,0.001939219,0.0004466093,0.001623085,0.004885363,0.001860163,0.000005539423],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006688273,"about_ca_system_score_gemma":0.001040702,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001522305,"about_ca_topic_score_gemma":0.000003676167,"domain_scores_codex":[0.9934077,0.000319002,0.0009458036,0.002950802,0.00121493,0.001161728],"domain_scores_gemma":[0.9896861,0.0007783944,0.001289267,0.007391749,0.0005605331,0.0002940124],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001192315,0.001239001,0.1640325,0.006237525,0.002240691,0.0002249194,0.0004310595,0.001127778,0.6083468,0.1239176,0.08875854,0.00225117],"study_design_scores_gemma":[0.002277643,0.0004380153,0.1231621,0.0008605422,0.0002644477,9.96801e-7,0.00002727978,0.5502915,0.1241885,0.00004069422,0.1950617,0.003386616],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08745418,0.003074148,0.8840716,0.011994,0.00264329,0.004684937,0.003329129,0.00266977,0.00007895372],"genre_scores_gemma":[0.8355865,0.001369932,0.159154,0.002177474,0.0003596203,0.00113393,0.0000181275,0.0001747631,0.00002563333],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7481323,"threshold_uncertainty_score":0.9996664,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04476784390055522,"score_gpt":0.2714744256389494,"score_spread":0.2267065817383942,"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."}}