{"id":"W2971290973","doi":"10.14778/3342263.3342633","title":"An intermediate representation for optimizing machine learning pipelines","year":2019,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":48,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Berlin Center for Machine Learning; Banting and Best Diabetes Centre, University of Toronto; York University","keywords":"Computer science; Preprocessor; Pipeline transport; Domain (mathematical analysis); External Data Representation; Representation (politics); Semantics (computer science); Data pre-processing; Feature (linguistics); Programming language; Feature engineering; Artificial intelligence; Theoretical computer science; Machine learning; Deep learning","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":[],"consensus_categories":[],"category_scores_codex":[0.00365664,0.0001119717,0.0002019945,0.0001932224,0.0001646316,0.0003722907,0.001320529,0.00002394864,0.00007444027],"category_scores_gemma":[0.001682123,0.00006789742,0.0001299249,0.0005110878,0.00004772451,0.0004444344,0.0005461128,0.0000910477,0.00003959761],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003003114,"about_ca_system_score_gemma":0.00001153751,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003893317,"about_ca_topic_score_gemma":0.000003193029,"domain_scores_codex":[0.9977226,0.00002133022,0.00054124,0.0005845365,0.0009160785,0.0002141993],"domain_scores_gemma":[0.9983515,0.0002851158,0.0005112295,0.000381907,0.0004148407,0.00005544929],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000588197,0.0005940286,0.3513296,0.0002440423,0.0001443313,5.251034e-7,0.009255592,0.03971445,0.2042449,0.0154891,0.05107969,0.3273155],"study_design_scores_gemma":[0.002221947,0.0005329134,0.01604081,0.0001879193,0.0000736933,0.000005961419,0.01083745,0.800462,0.1014142,0.02178699,0.04600423,0.000431851],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9877953,0.00004826567,0.004278206,0.001004265,0.001721616,0.0009229115,0.00001534965,0.00007066769,0.004143419],"genre_scores_gemma":[0.9900153,0.000006126225,0.005741078,0.00008536051,0.00008941763,0.00002317635,0.000007208806,0.000009809999,0.004022567],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7607476,"threshold_uncertainty_score":0.3590008,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08481634541627903,"score_gpt":0.3758174011326433,"score_spread":0.2910010557163643,"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."}}