{"id":"W7027101180","doi":"","title":"Big Data Clustering: Models and Applications","year":2023,"lang":"en","type":"dissertation","venue":"MacSphere (McMaster University)","topic":"Advanced Clustering Algorithms Research","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Mitacs","keywords":"Cluster analysis; Heuristic; Categorization; Big data; Identification (biology); Conceptual clustering; Similarity (geometry)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0001352107,0.0003161081,0.0002977148,0.0003791068,0.0002844494,0.0002713367,0.003529917,0.0002415243,0.0003577766],"category_scores_gemma":[0.00001219877,0.0003803128,0.00004936274,0.001191342,0.00005876787,0.0009357193,0.002999419,0.0004458251,0.0001165572],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001249225,"about_ca_system_score_gemma":0.0001699474,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001322243,"about_ca_topic_score_gemma":0.00115167,"domain_scores_codex":[0.9976146,0.00006762889,0.0001994628,0.001241098,0.0004227227,0.0004544489],"domain_scores_gemma":[0.9976975,0.0001059423,0.000143296,0.00171965,0.0001253396,0.0002082113],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00001885959,0.00001942899,0.00000390577,0.0001867092,0.00005046875,0.00007577927,0.0003422095,0.001208978,0.00002945683,0.001831905,0.0002840306,0.9959483],"study_design_scores_gemma":[0.0006576063,0.00005572108,0.0001257193,0.000173333,0.00005053001,0.00001590584,0.001903383,0.4738439,0.00003934234,0.003000929,0.5193919,0.0007417251],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"other","genre_scores_codex":[0.00002991815,0.000176446,0.8300162,0.00008585288,0.0004588044,0.0004866326,0.00006821998,0.0004763217,0.1682016],"genre_scores_gemma":[0.0005014151,0.0004564499,0.02539443,0.00003105657,0.0002495265,0.00001181558,0.000602139,0.00008187041,0.9726713],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.9952065,"threshold_uncertainty_score":0.9998649,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07643698189621714,"score_gpt":0.2879257419409178,"score_spread":0.2114887600447007,"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."}}