{"id":"W2073825221","doi":"10.1007/s10115-007-0090-6","title":"Robust projected clustering","year":2007,"lang":"en","type":"article","venue":"Knowledge and Information Systems","topic":"Advanced Statistical Methods and Models","field":"Mathematics","cited_by":70,"is_retracted":false,"has_abstract":false,"ca_institutions":"Simon Fraser University; University of Alberta","funders":"Yale University","keywords":"Cluster analysis; Categorical variable; Disjoint sets; Linear subspace; Outlier; Data mining; CURE data clustering algorithm; Cluster (spacecraft); Subspace topology; Set (abstract data type); Computer science; Data set; Correlation clustering; Clustering high-dimensional data; Single-linkage clustering; Algorithm; Pattern recognition (psychology); Mathematics; Artificial intelligence; Machine learning; Combinatorics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008255133,0.00009139613,0.0001604401,0.0001005866,0.0000920442,0.00005518978,0.00004614014,0.00006430932,0.000005295443],"category_scores_gemma":[0.0003200011,0.00007347397,0.00001790527,0.0001215545,0.00002136575,0.0008334933,0.00003622701,0.00007563677,0.00003353994],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003000185,"about_ca_system_score_gemma":0.00001490826,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000588111,"about_ca_topic_score_gemma":0.000008509334,"domain_scores_codex":[0.9991851,0.000031124,0.0004419923,0.00006978751,0.00009700502,0.0001750225],"domain_scores_gemma":[0.9992146,0.0003182588,0.0001140823,0.0001098408,0.0001642522,0.00007892845],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00004090311,0.00003689801,0.00007536812,0.001233038,0.00002332259,0.000001584472,0.00665426,0.00009137197,0.00009727305,0.8450027,0.00233829,0.1444051],"study_design_scores_gemma":[0.002441571,0.000256192,0.001112861,0.0006787335,0.00006310017,0.000175044,0.006964,0.3518253,0.0005484116,0.02908387,0.6059033,0.0009476296],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001789854,0.0001192062,0.9061499,0.000006040167,0.0002800014,0.0002830065,0.000007098544,0.0000863159,0.09127863],"genre_scores_gemma":[0.6925923,0.00004283055,0.3051424,0.00006464681,0.0002955481,0.0000502544,0.00002045289,0.00002148094,0.001770024],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8159187,"threshold_uncertainty_score":0.2996181,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1299341584742713,"score_gpt":0.3893189388673293,"score_spread":0.259384780393058,"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."}}