{"id":"W2912745458","doi":"10.1137/19m130858x","title":"Dual Space Preconditioning for Gradient Descent","year":2021,"lang":"en","type":"preprint","venue":"SIAM Journal on Optimization","topic":"Sparse and Compressive Sensing Techniques","field":"Engineering","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Army Research Laboratory; Army Research Office; Engineering and Physical Sciences Research Council; Ministry of Defence; Natural Sciences and Engineering Research Council of Canada; European Commission; Institute for Advanced Studies in the Humanities, University of Edinburgh; DeepMind; Seventh Framework Programme","keywords":"Convexity; Lipschitz continuity; Mathematics; Smoothness; Stochastic gradient descent; Gradient descent; Applied mathematics; Convex function; Gradient method; Nonlinear conjugate gradient method; Bregman divergence; Norm (philosophy); Proximal Gradient Methods; Mathematical analysis; Mathematical optimization; Regular polygon; Computer science; Geometry","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0001535194,0.0002526944,0.0002715463,0.0002006573,0.0001505658,0.0003726045,0.0001168963,0.0002231735,0.00006988662],"category_scores_gemma":[0.0000537666,0.0002705171,0.0001892719,0.00007567165,0.00001417799,0.000115736,0.00006134237,0.0006377304,0.000001782361],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000275879,"about_ca_system_score_gemma":0.00006255465,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000001343762,"about_ca_topic_score_gemma":0.000001455653,"domain_scores_codex":[0.9989542,0.00004687488,0.0003401634,0.0002230153,0.0001997514,0.0002360035],"domain_scores_gemma":[0.9991845,0.0000446232,0.0001839709,0.0002284832,0.0002626692,0.00009575567],"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":[0.00001285496,0.00003069582,0.000006109976,0.00004755582,0.0000935091,0.00001663523,0.0001462938,0.9932303,0.0001590676,0.0001420425,0.00481707,0.001297794],"study_design_scores_gemma":[0.0003284377,0.0001028331,0.00006787229,0.00141148,0.00008118134,0.000138714,0.00007257696,0.9808723,0.01454628,0.0008823182,0.001102409,0.000393625],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01690576,0.0006677106,0.977912,0.0001734928,0.002423841,0.0003712787,0.00001693067,0.0003881127,0.001140862],"genre_scores_gemma":[0.692149,0.002389484,0.3033514,0.0001201647,0.001236959,0.0000662666,0.0004104995,0.0001490607,0.0001272058],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6752432,"threshold_uncertainty_score":0.9999747,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02090318359265479,"score_gpt":0.2435101790497992,"score_spread":0.2226069954571444,"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."}}