{"id":"W3048804154","doi":"10.3390/electronics9081295","title":"The k-means Algorithm: A Comprehensive Survey and Performance Evaluation","year":2020,"lang":"en","type":"article","venue":"Electronics","topic":"Anomaly Detection Techniques and Applications","field":"Computer Science","cited_by":1609,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Cluster analysis; Computer science; Data mining; Algorithm; Convergence (economics); Initialization; Centroid; Outlier; Popularity; Variety (cybernetics); k-means clustering; Machine learning; Artificial intelligence","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":[0.003082341,0.002248439,0.003158832,0.00577372,0.001567727,0.003383148,0.002857314,0.002678999,0.002609636],"category_scores_gemma":[0.01158302,0.001150646,0.00206271,0.01144695,0.001024954,0.003923103,0.001400179,0.001907386,0.003607834],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001594711,"about_ca_system_score_gemma":0.002895416,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009061317,"about_ca_topic_score_gemma":0.005586541,"domain_scores_codex":[0.9956372,0.0006872825,0.0005225918,0.0008672286,0.002120981,0.000164729],"domain_scores_gemma":[0.9942123,0.002142913,0.0003178877,0.0004324975,0.002752179,0.0001422213],"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":[0.0003107737,0.0001243429,0.003343157,0.0030879,0.0003242073,0.00007672231,0.0002134342,0.04047912,0.002551206,0.006102369,0.0197121,0.9236748],"study_design_scores_gemma":[0.00009905621,0.001104712,0.01375121,0.002564521,0.0008560379,0.002374204,0.001197224,0.5639298,0.03291224,0.03499087,0.345567,0.0006531665],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"review","genre_scores_codex":[0.01918132,0.2661678,0.690882,0.001785933,0.001147579,0.000316563,0.001168913,0.004677642,0.01467224],"genre_scores_gemma":[0.1284844,0.2421637,0.6139997,0.0007910052,0.001331712,0.0004703496,0.004171844,0.001533423,0.007053891],"genre_candidate":"review","genre_consensus":null,"teacher_disagreement_score":0.009061317,"threshold_uncertainty_score":0.01801711,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02951616626777782,"score_gpt":0.2715012711448277,"score_spread":0.2419851048770499,"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."}}