{"id":"W4313495625","doi":"10.31449/inf.v46i4.3476","title":"OMCOKE: A Machine Learning Outlier-based Overlapping Clustering Technique for Multi-Label Data Analysis","year":2022,"lang":"en","type":"article","venue":"Informatica","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Fanshawe College","funders":"","keywords":"Cluster analysis; Outlier; Computer science; Anomaly detection; Data mining; Artificial intelligence; Unsupervised learning; Cluster (spacecraft); CURE data clustering algorithm; Machine learning; Pattern recognition (psychology); Correlation clustering","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":[0.003666594,0.001700017,0.002212404,0.005173019,0.002474859,0.00203149,0.003123415,0.002310688,0.001463204],"category_scores_gemma":[0.01387932,0.0006111047,0.002288383,0.006134634,0.001356732,0.002792394,0.003123215,0.003194888,0.00126299],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001308742,"about_ca_system_score_gemma":0.002433878,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006416671,"about_ca_topic_score_gemma":0.00968244,"domain_scores_codex":[0.9950359,0.001055245,0.0004063966,0.001290105,0.00187504,0.0003373408],"domain_scores_gemma":[0.9935296,0.002071674,0.001015821,0.001213485,0.001962703,0.0002066495],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0008094424,0.0005270476,0.01057865,0.0007358391,0.0008681375,0.000560323,0.001607437,0.1368389,0.01702804,0.01792519,0.01694467,0.7955762],"study_design_scores_gemma":[0.00006020974,0.0001294737,0.003223891,0.00008700452,0.00008904246,0.0004729251,0.0004503583,0.9419409,0.01299537,0.02686146,0.01356629,0.0001231273],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.006718704,0.0002461928,0.9897476,0.0001292226,0.00007155966,0.0001418813,0.0002104315,0.002277392,0.0004570169],"genre_scores_gemma":[0.1052391,0.0002115221,0.8908994,0.0001430278,0.00008341104,0.0003953573,0.001233596,0.0004845087,0.001310198],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006416671,"threshold_uncertainty_score":0.01939106,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09572787538383581,"score_gpt":0.3254020660487998,"score_spread":0.229674190664964,"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."}}