{"id":"W2096830858","doi":"10.1007/s00224-014-9539-7","title":"Probabilistic k-Median Clustering in Data Streams","year":2014,"lang":"en","type":"article","venue":"Theory of Computing Systems","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":false,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Probabilistic logic; Mathematics; Cluster analysis; Representation (politics); Metric (unit); Probability distribution; Euclidean geometry; Point (geometry); Probabilistic analysis of algorithms; Combinatorics; Statistics","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":[0.005619589,0.0007239535,0.002510605,0.002597556,0.001394734,0.002937029,0.003353501,0.002251491,0.001571546],"category_scores_gemma":[0.0246875,0.001154367,0.001394384,0.004352227,0.001561441,0.004481504,0.002188617,0.001874874,0.0005373481],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002183562,"about_ca_system_score_gemma":0.001547537,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004731608,"about_ca_topic_score_gemma":0.003773918,"domain_scores_codex":[0.9964414,0.001128892,0.000315441,0.0007546116,0.001068683,0.0002908161],"domain_scores_gemma":[0.9883744,0.007418655,0.0009569821,0.001120878,0.001796031,0.0003331091],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005072544,0.00009071727,0.002881279,0.0002228127,0.0001171029,0.0001099933,0.0002718801,0.8565718,0.00137186,0.06582125,0.002330668,0.06970332],"study_design_scores_gemma":[0.000007724105,0.00001198798,0.0002168287,0.000007941034,0.000006783321,0.00002320094,0.00002807562,0.9679255,0.0003042706,0.03106951,0.0003891258,0.000008908764],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01593732,0.0003736163,0.9826132,0.0002474271,0.00005388169,0.00003716746,0.0001186883,0.0002399994,0.0003787543],"genre_scores_gemma":[0.6282867,0.001217302,0.3640312,0.0001562156,0.0003998935,0.0002812068,0.001071751,0.0002552545,0.004300567],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005619589,"threshold_uncertainty_score":0.02971959,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03428720038795635,"score_gpt":0.2571392949994608,"score_spread":0.2228520946115044,"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."}}