{"id":"W4407295943","doi":"10.1002/sim.70013","title":"Challenges for Predictive Modeling With Neural Network Techniques Using Error‐Prone Dietary Intake Data","year":2025,"lang":"en","type":"article","venue":"Statistics in Medicine","topic":"Nutritional Studies and Diet","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; University of New Brunswick","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Overfitting; Computer science; Machine learning; Artificial neural network; Artificial intelligence; Replicate; Predictive modelling; Data mining; Statistics; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.05630674,0.001703603,0.002125859,0.002152742,0.001201087,0.004954447,0.004427749,0.002348889,0.001490123],"category_scores_gemma":[0.1604586,0.0014617,0.001579819,0.002506173,0.003306751,0.005110102,0.004153357,0.008122899,0.0007361829],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002516702,"about_ca_system_score_gemma":0.00340324,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01946017,"about_ca_topic_score_gemma":0.01358462,"domain_scores_codex":[0.9810126,0.01403597,0.001021469,0.001584721,0.002089487,0.0002557519],"domain_scores_gemma":[0.8408537,0.143262,0.003915398,0.00671547,0.004820786,0.000432655],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001292481,0.0001152886,0.01209832,0.0005583039,0.0006488986,0.0002214426,0.0005601564,0.7686666,0.0006113863,0.07648002,0.003314444,0.1365959],"study_design_scores_gemma":[0.00001400434,0.00002816029,0.000906478,0.000127389,0.00002645273,0.00004031679,0.00007411306,0.9104692,0.0002889033,0.08601858,0.001975331,0.00003107322],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01303594,0.002539214,0.9740249,0.00771871,0.0001631794,0.0001227456,0.0002725875,0.000417722,0.001704953],"genre_scores_gemma":[0.4587928,0.005503373,0.5276871,0.002496757,0.0008357848,0.0009014811,0.0008148181,0.0003169078,0.002650955],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.05630674,"threshold_uncertainty_score":0.2977821,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1615857319248665,"score_gpt":0.4038552868871068,"score_spread":0.2422695549622403,"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."}}