{"id":"W3174585579","doi":"10.1609/aaai.v36i4.20284","title":"The Perils of Learning Before Optimizing","year":2022,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Transportation Planning and Optimization","field":"Social Sciences","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada; Alberta Machine Intelligence Institute; Compute Canada; Defense Advanced Research Projects Agency; Canadian Institute for Advanced Research; Nvidia","keywords":"A priori and a posteriori; Computer science; Range (aeronautics); Function (biology); Task (project management); Mathematical optimization; Optimization problem; Machine learning; Artificial intelligence; Algorithm; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":["sts"],"consensus_categories":[],"category_scores_codex":[0.001145938,0.00008377044,0.0001214578,0.00005801231,0.00183678,0.00008264647,0.0007002086,0.00003472094,0.0002019213],"category_scores_gemma":[0.0004937595,0.00006074976,0.00008031118,0.0005434714,0.0004349295,0.0001196297,0.00006211825,0.000299724,0.000003950067],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004232855,"about_ca_system_score_gemma":0.0001383957,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002120595,"about_ca_topic_score_gemma":0.00008768214,"domain_scores_codex":[0.9986154,0.000049928,0.000358164,0.0001645562,0.0006005429,0.0002114336],"domain_scores_gemma":[0.9989437,0.0001243275,0.0004041702,0.00007647802,0.0004128781,0.00003847411],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"qualitative","study_design_scores_codex":[0.00008974365,0.00004775951,0.002486948,0.00001159821,0.00001153344,1.161441e-7,0.04036379,0.01572158,0.002298465,0.9175592,0.0000711464,0.02133813],"study_design_scores_gemma":[0.000134805,0.001192748,0.002646531,0.0003967487,0.0001136498,0.000003093629,0.6150132,0.07351031,0.127654,0.1573038,0.02130077,0.0007303378],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9054127,0.00008943012,0.002101881,0.01074848,0.001030851,0.0007941841,0.00001724456,0.0001440923,0.07966114],"genre_scores_gemma":[0.9982531,0.00005343405,0.0003768577,0.00003216407,0.00003870515,0.00002467974,0.000001196471,0.000007566062,0.001212274],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7602553,"threshold_uncertainty_score":0.9994627,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05691481360200981,"score_gpt":0.3112706199021111,"score_spread":0.2543558063001014,"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."}}