{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007334708,0.001634337,0.001967716,0.004614292,0.001640867,0.006082701,0.004087827,0.003652181,0.003931646],"category_scores_gemma":[0.02190646,0.001273888,0.002376051,0.008384294,0.003072859,0.007477703,0.003888231,0.003325692,0.001633206],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003800136,"about_ca_system_score_gemma":0.00209126,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006829762,"about_ca_topic_score_gemma":0.004718963,"domain_scores_codex":[0.9953446,0.002055481,0.0003089916,0.0009459185,0.001165252,0.0001797572],"domain_scores_gemma":[0.9861385,0.009593642,0.0009357184,0.001524066,0.001516943,0.0002911154],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00004408601,0.00008702325,0.003113011,0.0006337331,0.0002032962,0.0001486991,0.0004915947,0.2090581,0.0002068564,0.6730668,0.01506514,0.09788172],"study_design_scores_gemma":[0.00001010544,0.00001642719,0.0005099151,0.0001257505,0.00002157781,0.00008730518,0.000137703,0.5030043,0.0001448389,0.4804855,0.01542795,0.00002858901],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0043314,0.008882922,0.9716918,0.004946658,0.000370631,0.0002573353,0.0008895714,0.0007133276,0.007916315],"genre_scores_gemma":[0.232914,0.02143192,0.7277683,0.001458824,0.002110317,0.001763112,0.003075117,0.0003926677,0.009085883],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007334708,"threshold_uncertainty_score":0.03879011,"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."}}