{"id":"W2140511449","doi":"10.1007/978-3-540-24840-8_37","title":"Distributed Data Mining vs. Sampling Techniques: A Comparison","year":2004,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université Laval","funders":"","keywords":"Computer science; Data mining; Download; Aggregate (composite); Sampling (signal processing); Distributed database; Set (abstract data type); Volume (thermodynamics); Database; World Wide Web","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":[0.01189661,0.001187889,0.003367195,0.004035045,0.0007454573,0.002624379,0.003637281,0.001836799,0.004305345],"category_scores_gemma":[0.03387636,0.000633489,0.001538745,0.005981235,0.0008989126,0.00491741,0.001749856,0.001572877,0.001287922],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001336495,"about_ca_system_score_gemma":0.001818643,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001934803,"about_ca_topic_score_gemma":0.003622814,"domain_scores_codex":[0.9904888,0.00360014,0.0005191086,0.001012117,0.004084121,0.0002958079],"domain_scores_gemma":[0.9483685,0.03908953,0.001121194,0.005256515,0.00533995,0.0008243406],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.003279387,0.0005141418,0.005910326,0.001872547,0.0006967035,0.00005221879,0.000238646,0.01498032,0.002866155,0.01113992,0.00545691,0.9529927],"study_design_scores_gemma":[0.002849925,0.006186087,0.0364757,0.001569163,0.002841303,0.002864637,0.002080663,0.7391214,0.02042149,0.1333896,0.05195807,0.0002419753],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.09874184,0.0726612,0.8017805,0.00349973,0.001333715,0.0009914399,0.0007700839,0.00386356,0.01635797],"genre_scores_gemma":[0.5125848,0.03688411,0.4392339,0.001091808,0.001746178,0.0006004991,0.001383286,0.0009730632,0.005502355],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01189661,"threshold_uncertainty_score":0.06291604,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06064199965264973,"score_gpt":0.3188307178509959,"score_spread":0.2581887181983462,"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."}}