{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00200847,0.0008986053,0.001860843,0.0005914084,0.00105826,0.001520215,0.003904176,0.001561925,0.003182174],"category_scores_gemma":[0.005247181,0.000675851,0.000990986,0.001097981,0.0005444711,0.003694092,0.002691079,0.002705659,0.001638658],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007606695,"about_ca_system_score_gemma":0.001878547,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005877874,"about_ca_topic_score_gemma":0.0122133,"domain_scores_codex":[0.9988305,0.0002542454,0.00005976758,0.0003476012,0.0003697744,0.0001380751],"domain_scores_gemma":[0.9974636,0.000678102,0.0001021047,0.0009112055,0.000697122,0.0001477547],"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":[0.0005932327,0.0005234783,0.002729523,0.0000774855,0.0001730104,0.0001373586,0.0001586918,0.5143271,0.007383924,0.009248092,0.008701949,0.4559462],"study_design_scores_gemma":[0.000008284867,0.0000203189,0.00007985581,0.00000181692,0.00001156286,0.00001374112,0.000007583071,0.9949493,0.0006333975,0.003891747,0.0003780884,0.000004185889],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009580781,0.0001156764,0.9878194,0.000143673,0.0000706374,0.00005822728,0.00008269201,0.001491165,0.000637796],"genre_scores_gemma":[0.448376,0.0002421932,0.542465,0.0003143918,0.0003056724,0.000380343,0.0007679663,0.0005217772,0.00662671],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005877874,"threshold_uncertainty_score":0.01168734,"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."}}