{"id":"W2990170152","doi":"10.2196/13347","title":"Lifestyle Disease Surveillance Using Population Search Behavior: Feasibility Study","year":2019,"lang":"en","type":"article","venue":"Journal of Medical Internet Research","topic":"Data-Driven Disease Surveillance","field":"Medicine","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Overfitting; Generalizability theory; Multivariate statistics; Covariate; Computer science; Context (archaeology); Proxy (statistics); Population; Feature selection; Medicine; Statistics; Machine learning; Geography; Environmental health; Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.01446136,0.0001828612,0.0006681805,0.0004889134,0.00005849686,0.00009763142,0.0008143352,0.0001496131,0.003438202],"category_scores_gemma":[0.00482202,0.0001394029,0.0002375175,0.000579089,0.000222646,0.0002481578,0.0005110182,0.00193438,0.0001890516],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006146473,"about_ca_system_score_gemma":0.001191864,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008295259,"about_ca_topic_score_gemma":0.000100181,"domain_scores_codex":[0.9889358,0.001800953,0.0009966887,0.0004502755,0.007215266,0.0006009633],"domain_scores_gemma":[0.9952386,0.0005819955,0.0002094393,0.0008189598,0.001422288,0.001728735],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001810789,0.002736842,0.9876783,0.0001776523,0.0001004914,0.001930366,0.0001430882,0.00001252017,0.0001905469,0.00002065565,0.002539951,0.002658743],"study_design_scores_gemma":[0.003061177,0.001198788,0.9878259,0.0004684218,0.00004278729,0.00020668,0.0004521113,0.006101747,0.00002256927,0.00002014499,0.0004803907,0.0001192655],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9965965,0.000470502,0.00006509636,0.000983925,0.0003719854,0.001165473,0.00002749409,0.00002508905,0.0002939657],"genre_scores_gemma":[0.9984,0.00007021436,0.0000917763,0.00007184171,0.000562829,0.00000730084,0.00003468763,0.00003547799,0.0007259179],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009639343,"threshold_uncertainty_score":0.9974728,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1238580738224365,"score_gpt":0.4834293947039309,"score_spread":0.3595713208814943,"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."}}