{"id":"W2068406209","doi":"10.1016/j.patcog.2008.06.022","title":"Discrete data clustering using finite mixture models","year":2008,"lang":"en","type":"article","venue":"Pattern Recognition","topic":"Bayesian Methods and Mixture Models","field":"Computer Science","cited_by":48,"is_retracted":false,"has_abstract":false,"ca_institutions":"Concordia University","funders":"","keywords":"Mixture model; Automatic summarization; Computer science; Robustness (evolution); Cluster analysis; Pattern recognition (psychology); Artificial intelligence; Dirichlet distribution; Algorithm; Data mining; Mathematics","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":[0.005012462,0.001399564,0.003704531,0.004820412,0.001352537,0.00410388,0.005370632,0.002955835,0.002256021],"category_scores_gemma":[0.01987227,0.002056445,0.003611563,0.004899302,0.002831479,0.005226979,0.003067474,0.003855811,0.001850406],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002130843,"about_ca_system_score_gemma":0.001325523,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007080008,"about_ca_topic_score_gemma":0.00663222,"domain_scores_codex":[0.9943742,0.002448295,0.000351179,0.001229367,0.001387004,0.0002099412],"domain_scores_gemma":[0.9889123,0.007105567,0.0006694774,0.001926644,0.001173161,0.0002127717],"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":[0.0001856532,0.0001109137,0.001402738,0.0004115246,0.0003828242,0.00009445448,0.000362505,0.5873293,0.003490208,0.1358865,0.003421745,0.2669215],"study_design_scores_gemma":[0.000008991689,0.000008980282,0.0001834659,0.00002474658,0.00002190164,0.00005436493,0.0000186985,0.9220655,0.0008947248,0.07555911,0.001130087,0.00002944797],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0006981959,0.000206382,0.9987609,0.00003892961,0.00001088221,0.00001051691,0.00001881775,0.0001367279,0.000118652],"genre_scores_gemma":[0.07930121,0.0009272649,0.916807,0.0001061913,0.0001026469,0.0002300065,0.000510104,0.0002728451,0.0017428],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007080008,"threshold_uncertainty_score":0.02650881,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1837551475213755,"score_gpt":0.3141076386796353,"score_spread":0.1303524911582598,"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."}}