{"id":"W4409120969","doi":"10.1007/s10898-025-01483-8","title":"Support vector machines with the hard-margin loss: optimal training via combinatorial Benders’ cuts","year":2025,"lang":"en","type":"article","venue":"Journal of Global Optimization","topic":"Face and Expression Recognition","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"Polytechnique Montréal","funders":"Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro; Conselho Nacional de Desenvolvimento Científico e Tecnológico; Coordenação de Aperfeiçoamento de Pessoal de Nível Superior","keywords":"Mathematics; Margin (machine learning); Mathematical optimization; Benders' decomposition; Training (meteorology); Combinatorial optimization; Support vector machine; Combinatorics; Applied mathematics; Algorithm; Artificial intelligence; Machine learning; Computer science","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":[0.003636762,0.002342878,0.003551041,0.001224762,0.0006425168,0.002253686,0.002977846,0.004575834,0.007804316],"category_scores_gemma":[0.01692909,0.002181326,0.001149942,0.001656206,0.001907045,0.00456762,0.002624888,0.005897358,0.002210401],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001262797,"about_ca_system_score_gemma":0.001677422,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002002443,"about_ca_topic_score_gemma":0.001918449,"domain_scores_codex":[0.9978445,0.001155299,0.00009856022,0.0003476056,0.0003694985,0.0001845028],"domain_scores_gemma":[0.9935482,0.00531301,0.0002652303,0.0003336182,0.0003492708,0.000190732],"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":[0.0005387354,0.0002901304,0.0004127152,0.0002515285,0.00007266431,0.00007519126,0.00008488452,0.78045,0.001515237,0.04607132,0.01115985,0.1590777],"study_design_scores_gemma":[0.00003021374,0.00004256438,0.00004058026,0.00001634751,0.00000455681,0.00001153451,0.000007768697,0.9773639,0.0001857396,0.02199861,0.0002930991,0.000005069024],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009884452,0.0005194134,0.9854833,0.000677105,0.00006436365,0.00008114731,0.0001122531,0.0005968232,0.00258112],"genre_scores_gemma":[0.333528,0.0007514528,0.6523615,0.0008831125,0.0003990824,0.0007984419,0.001196554,0.0007746686,0.009307242],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007804316,"threshold_uncertainty_score":0.02610803,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01007542065080586,"score_gpt":0.2427509717798532,"score_spread":0.2326755511290473,"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."}}