{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007656953,0.001254004,0.001285677,0.003716325,0.001115543,0.001543108,0.003167917,0.001787899,0.002964204],"category_scores_gemma":[0.02504475,0.0006835017,0.001516449,0.002962257,0.001166618,0.002675663,0.002003735,0.002682919,0.002164973],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001155095,"about_ca_system_score_gemma":0.002106611,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003088297,"about_ca_topic_score_gemma":0.003561451,"domain_scores_codex":[0.9953871,0.0018513,0.0002040688,0.001301197,0.001085937,0.0001702764],"domain_scores_gemma":[0.9903285,0.004913658,0.001095014,0.001718989,0.001731358,0.0002124646],"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":[0.0003881694,0.0003025928,0.009003283,0.0006399544,0.0005275291,0.0002274063,0.000667445,0.3174671,0.01770675,0.06044024,0.01293978,0.5796897],"study_design_scores_gemma":[0.00002718879,0.00004161197,0.001095772,0.00002927008,0.00002585989,0.00008526577,0.00004632898,0.9699186,0.004513799,0.02057144,0.003601739,0.00004307317],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002457638,0.00006845218,0.9963906,0.0000720849,0.00001604793,0.00004429622,0.00009261297,0.0006588412,0.0001993264],"genre_scores_gemma":[0.07048367,0.0001624561,0.9264068,0.0001160654,0.00005734263,0.0003569403,0.0006505605,0.0003208029,0.001445343],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007656953,"threshold_uncertainty_score":0.04049426,"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."}}