{"id":"W4390244867","doi":"10.18280/ria.370606","title":"Evaluating Binary Classification Algorithms on Data Lakes Using Machine Learning","year":2023,"lang":"en","type":"article","venue":"Revue d intelligence artificielle","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Department of Artificial Intelligence, Korea University; Lviv Polytechnic National University; Fonds National de la Recherche Luxembourg","keywords":"Computer science; Binary number; Binary classification; Machine learning; Artificial intelligence; Binary data; Data mining; Support vector machine; Mathematics; Arithmetic","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"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":[0.006797967,0.001039368,0.001220183,0.002922476,0.000961971,0.002465104,0.001361571,0.001842246,0.001681043],"category_scores_gemma":[0.02785073,0.0003549987,0.0008713042,0.002048644,0.0008472767,0.003495669,0.001306507,0.001028163,0.0004123578],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001821651,"about_ca_system_score_gemma":0.001414156,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006389289,"about_ca_topic_score_gemma":0.005003267,"domain_scores_codex":[0.9967325,0.001349114,0.0003273549,0.0004532498,0.0009555019,0.0001822895],"domain_scores_gemma":[0.9643463,0.02907349,0.0009328459,0.001107936,0.004082635,0.0004567808],"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":[0.003590621,0.001349188,0.04102163,0.0005172311,0.000469462,0.0001196985,0.0002189892,0.5397522,0.006716689,0.007830959,0.005417838,0.3929956],"study_design_scores_gemma":[0.00003680744,0.0001668612,0.0009407466,0.000008330727,0.00002378147,0.00001025117,0.00004083762,0.9953082,0.002017768,0.001299563,0.0001423494,0.000004473006],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8284211,0.001319739,0.1652059,0.0006302345,0.0003114792,0.0002278652,0.0003504418,0.001271011,0.002262253],"genre_scores_gemma":[0.8943479,0.0002765229,0.10252,0.00008843709,0.00009456903,0.0001317074,0.0008720802,0.0001030158,0.001565656],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006797967,"threshold_uncertainty_score":0.03595155,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3618588023830966,"score_gpt":0.4242256351558663,"score_spread":0.06236683277276961,"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."}}