{"id":"W7104458993","doi":"10.71781/15656","title":"On the generalization of machine learning models in finance : five essays on bridging the empirical gap","year":2025,"lang":"en","type":"dissertation","venue":"Open MIND","topic":"Stock Market Forecasting Methods","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Mitacs","keywords":"Repartition; Context (archaeology); Decision tree","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.008491664,0.0002375249,0.0004592714,0.0003296834,0.0002912624,0.0003228336,0.001939646,0.0001734239,0.0005698279],"category_scores_gemma":[0.01405758,0.0001298248,0.0001166777,0.001276073,0.00004196848,0.0001362928,0.0002181555,0.0006832314,0.00003720821],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000066239,"about_ca_system_score_gemma":0.00026103,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002365383,"about_ca_topic_score_gemma":0.0009314129,"domain_scores_codex":[0.9954013,0.001893588,0.0007751818,0.0006622848,0.00105618,0.0002115145],"domain_scores_gemma":[0.9894298,0.008841435,0.0007641684,0.000702524,0.0002413098,0.00002078356],"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":[0.0004572532,0.00006033411,0.0005106686,0.000007765748,0.00001746046,0.00000475618,0.0067369,0.3188931,0.00003779952,0.002657087,0.003607339,0.6670095],"study_design_scores_gemma":[0.0003796894,0.000130155,0.00444579,0.0008688672,0.00003121726,0.000001444224,0.001606736,0.9355248,0.001409599,0.05102647,0.004304118,0.0002711576],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7575312,0.0001387561,0.004691191,0.0008586762,0.0005314572,0.0009255022,0.00003173073,0.000002056863,0.2352894],"genre_scores_gemma":[0.8498309,0.00004047099,0.005166518,0.0002437466,0.0000707892,0.00009781707,0.0001590754,0.00003369674,0.1443569],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6667384,"threshold_uncertainty_score":0.9942474,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3533241481307882,"score_gpt":0.4694936314239501,"score_spread":0.116169483293162,"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."}}