{"id":"W2568188511","doi":"10.1016/j.jda.2011.10.002","title":"Indexability, concentration, and VC theory","year":2011,"lang":"en","type":"article","venue":"Journal of Discrete Algorithms","topic":"Bayesian Methods and Mixture Models","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Ottawa","funders":"","keywords":"Search engine indexing; Histogram; Measure (data warehouse); Dimension (graph theory); Similarity (geometry); Similarity measure; Statistical analysis","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":[0.007010133,0.001326116,0.002869253,0.004145369,0.002200799,0.006621569,0.0035137,0.002760191,0.009402183],"category_scores_gemma":[0.05994115,0.0009326347,0.001502464,0.004754498,0.007033778,0.01191952,0.006064988,0.004688805,0.0008460871],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003781259,"about_ca_system_score_gemma":0.001540854,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002823981,"about_ca_topic_score_gemma":0.001479403,"domain_scores_codex":[0.9952461,0.002307777,0.0002042914,0.0006525681,0.001218467,0.0003708414],"domain_scores_gemma":[0.9575852,0.03435864,0.001951458,0.002608459,0.002360646,0.001135557],"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":[0.00002716246,0.00001598578,0.0002993609,0.00004682675,0.00001744495,0.00001517289,0.00006124187,0.007516779,0.0001413935,0.9824737,0.001485644,0.007899277],"study_design_scores_gemma":[0.000009473041,0.00000878961,0.00009310077,0.00001888744,0.000007430662,0.00002261758,0.0000144617,0.04018482,0.0001275669,0.9580464,0.001456748,0.000009759096],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02954345,0.004054298,0.9309245,0.003680068,0.0002164842,0.00006107568,0.0002762892,0.0002225055,0.03102137],"genre_scores_gemma":[0.8104572,0.007285663,0.1477693,0.002036696,0.002133962,0.0005146967,0.001127559,0.0006593684,0.02801551],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009402183,"threshold_uncertainty_score":0.03707355,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02544690606395558,"score_gpt":0.2651457390474658,"score_spread":0.2396988329835102,"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."}}