{"id":"W2901153708","doi":"10.1016/j.bdr.2018.11.001","title":"Label-Aware Distributed Ensemble Learning: A Simplified Distributed Classifier Training Model for Big Data","year":2018,"lang":"en","type":"article","venue":"Big Data Research","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":23,"is_retracted":false,"has_abstract":false,"ca_institutions":"IBM (Canada); Queen's University","funders":"","keywords":"Computer science; Classifier (UML); Big data; Data mining; Machine learning; SPARK (programming language); Artificial intelligence","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":["metaepi_narrow","scholarly_communication","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.006154601,0.0003720792,0.0004610755,0.0004204068,0.00106911,0.001193128,0.01728046,0.0002862188,0.000008023857],"category_scores_gemma":[0.005558135,0.0003556625,0.00004436456,0.001953894,0.0005941198,0.001694246,0.01924838,0.0009477636,0.00009374799],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000167606,"about_ca_system_score_gemma":0.00127327,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000212479,"about_ca_topic_score_gemma":0.0004531641,"domain_scores_codex":[0.9933143,0.0004758724,0.0005982956,0.002464965,0.001417662,0.001728958],"domain_scores_gemma":[0.9867419,0.001244639,0.0001830588,0.01041267,0.0009835563,0.000434162],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001271442,0.0002382074,0.00009078832,0.00006287715,0.00007747657,0.00003270503,0.0004360279,0.00001731861,0.001498988,0.002379782,0.5699521,0.4250866],"study_design_scores_gemma":[0.0006749828,0.000287997,0.00005716614,0.00007163992,0.00001257898,0.00001340512,0.0001286383,0.7455682,0.000797228,0.002101301,0.2499517,0.0003351819],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0005186178,0.00005966037,0.9506995,0.001862755,0.0003666434,0.0007793141,0.04462689,0.0008155096,0.0002710951],"genre_scores_gemma":[0.7522509,0.00008633828,0.1183348,0.0001359976,0.001461657,0.0002330858,0.126477,0.0001037017,0.0009164897],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8323647,"threshold_uncertainty_score":0.9998896,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.72474036327115,"score_gpt":0.4627651775700344,"score_spread":0.2619751857011156,"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."}}