{"id":"W3002371932","doi":"10.1371/journal.pone.0227788","title":"Blind method for discovering number of clusters in multidimensional datasets by regression on linkage hierarchies generated from random data","year":2020,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Complex Network Analysis Techniques","field":"Physics and Astronomy","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"Faculty of Health Sciences, Queen's University","keywords":"Computer science; Outlier; Cluster analysis; Data mining; Hierarchical clustering; Regression; Regression analysis; Exploratory data analysis; Artificial intelligence; Pattern recognition (psychology); Statistics; Mathematics; Machine learning","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":[],"consensus_categories":[],"category_scores_codex":[0.0001873345,0.0001481563,0.0004259852,0.00003763418,0.00004598798,0.00001993693,0.000302188,0.00003187316,0.0001447245],"category_scores_gemma":[0.00003672718,0.0001277736,0.0000544309,0.0001635502,0.00002346636,0.0001348034,0.0003461248,0.0001612745,0.000003932182],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009537317,"about_ca_system_score_gemma":0.0000178494,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003177875,"about_ca_topic_score_gemma":0.00001685304,"domain_scores_codex":[0.9987682,0.0001335331,0.0003268167,0.0004114804,0.0002108509,0.0001491653],"domain_scores_gemma":[0.9988229,0.0004587206,0.0001515022,0.0004796105,0.00002520794,0.00006209331],"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":[0.006497702,0.005447017,0.03462,0.0001819861,0.002931272,0.000006103156,0.001134744,0.002370387,0.8588138,0.0007495675,0.06734488,0.01990253],"study_design_scores_gemma":[0.005385981,0.00005968938,0.000154795,0.0004503683,0.0002672128,3.022808e-8,0.00006721451,0.7779076,0.2139003,0.0005721059,0.000956976,0.0002777287],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8340243,0.00006734351,0.1538439,0.0004550276,0.00001218429,0.0005241908,0.01097294,0.00003337005,0.00006673898],"genre_scores_gemma":[0.6947474,0.00001206902,0.2716253,0.0001201942,0.0002174879,0.00004128873,0.03319053,0.00002597478,0.00001976075],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7755372,"threshold_uncertainty_score":0.5210456,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08603208803713068,"score_gpt":0.3368215724897964,"score_spread":0.2507894844526657,"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."}}