{"id":"W4319653775","doi":"10.48550/arxiv.2302.03391","title":"Sparse and geometry-aware generalisation of the mutual information for joint discriminative clustering and feature selection","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Advanced Clustering Algorithms Research","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre hospitalier de l'Université Laval","funders":"Agence Nationale de la Recherche","keywords":"Cluster analysis; Discriminative model; Feature selection; Pattern recognition (psychology); Computer science; Scalability; Selection (genetic algorithm); Feature (linguistics); Artificial intelligence; Relevance (law); Clustering high-dimensional data; Data mining; Machine learning; 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":[0.004245994,0.001810529,0.002321515,0.00192986,0.0009489047,0.001594109,0.003411211,0.002528845,0.002171012],"category_scores_gemma":[0.01346191,0.0009818607,0.001822422,0.002850422,0.0022585,0.002733807,0.003966189,0.002813865,0.001216058],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001608982,"about_ca_system_score_gemma":0.001257075,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002653602,"about_ca_topic_score_gemma":0.003924415,"domain_scores_codex":[0.9971545,0.001270895,0.0001156158,0.0006567638,0.0006400489,0.0001621937],"domain_scores_gemma":[0.9961786,0.002029273,0.0003604684,0.0007375841,0.0005002223,0.000193861],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0002365425,0.0001235115,0.001402934,0.0002251908,0.000154969,0.0001390925,0.000240533,0.7622107,0.008336188,0.05030099,0.005685898,0.1709434],"study_design_scores_gemma":[0.00001547671,0.00005368901,0.0003246931,0.000009885185,0.00001047175,0.00007658611,0.00001912625,0.973409,0.001330708,0.02356977,0.001163115,0.00001755679],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005052437,0.0001497127,0.9936186,0.0001763279,0.00001695982,0.0000448791,0.00006515513,0.0003768797,0.0004990608],"genre_scores_gemma":[0.2377904,0.0003838758,0.7563463,0.0004447426,0.0001891456,0.0003958406,0.001165397,0.0004369976,0.002847239],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004245994,"threshold_uncertainty_score":0.02245522,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1268804352697365,"score_gpt":0.2295491491202739,"score_spread":0.1026687138505374,"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."}}