{"id":"W2072065544","doi":"10.1145/1923947.1923983","title":"An introduction to data mining and predictive analytics","year":2010,"lang":"en","type":"article","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"IBM (Canada)","funders":"","keywords":"Computer science; Analytics; Data science; Process (computing); Data mining; Predictive analytics; Data analysis; 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":[],"consensus_categories":[],"category_scores_codex":[0.0002071563,0.00004240629,0.000042447,0.00004339819,0.00007688756,0.0001769739,0.0007638397,0.00001841764,0.00001113205],"category_scores_gemma":[0.00004428423,0.00003748367,0.000002800629,0.000202716,0.00001975013,0.0007416183,0.0003999872,0.00005265194,0.000006947388],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000002440908,"about_ca_system_score_gemma":0.00001522633,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000139334,"about_ca_topic_score_gemma":0.00004768984,"domain_scores_codex":[0.9993802,0.000005680495,0.00006832825,0.000389482,0.00007345028,0.0000828237],"domain_scores_gemma":[0.998542,0.00001779282,0.00001658711,0.001291143,0.00003514075,0.00009737589],"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":[0.000001557729,0.00007323201,0.0006470963,0.000001791272,0.00001069464,7.754717e-7,0.000687902,0.00002750082,0.006482872,0.03034173,0.08400758,0.8777173],"study_design_scores_gemma":[0.00003567453,0.00004973903,0.004232501,5.844177e-7,0.000003840191,0.000009102669,0.00007507055,0.9524111,0.0003273712,0.0001560314,0.04263629,0.00006276578],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0262164,0.000001331393,0.9670757,0.005963905,0.0001773793,0.00005588615,0.00004083532,0.00008222585,0.0003863049],"genre_scores_gemma":[0.1316031,9.913331e-7,0.8675951,0.0001623852,0.0004240829,0.000004593682,0.00007588244,0.000002694507,0.0001312374],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9523835,"threshold_uncertainty_score":0.1706564,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02519771867409121,"score_gpt":0.298142209585086,"score_spread":0.2729444909109948,"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."}}