{"id":"W4323035821","doi":"10.1099/mgen.0.000929","title":"Epidemiological cluster identification using multiple data sources: an approach using logistic regression","year":2023,"lang":"en","type":"article","venue":"Microbial Genomics","topic":"Data-Driven Disease Surveillance","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"European Commission","keywords":"Cluster analysis; Pairwise comparison; Spatial epidemiology; Cluster (spacecraft); Data mining; Computational biology; Epidemiology; Logistic regression; Biology; Computer science; Statistics; Mathematics; Medicine","routes":{"ca_aff":true,"ca_fund":false,"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":[0.0550755,0.002880214,0.004694788,0.0138802,0.002625763,0.006012991,0.009475983,0.003772911,0.002875927],"category_scores_gemma":[0.09309641,0.002254408,0.006492586,0.01409784,0.001717219,0.005449148,0.006882819,0.006111738,0.001662328],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002322807,"about_ca_system_score_gemma":0.003838594,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01559481,"about_ca_topic_score_gemma":0.009798156,"domain_scores_codex":[0.9516525,0.03419542,0.002276944,0.007116095,0.003761492,0.0009974781],"domain_scores_gemma":[0.9100702,0.07059736,0.006398388,0.006090767,0.005873053,0.0009702832],"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":[0.001322744,0.00131599,0.2034315,0.001257957,0.00794606,0.001740834,0.002991494,0.3195623,0.001512126,0.02698749,0.009280547,0.422651],"study_design_scores_gemma":[0.0000520322,0.00008421651,0.003995524,0.00009702666,0.0002184118,0.0001591214,0.000384625,0.9756821,0.0003362998,0.01661709,0.002292758,0.00008086095],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02247619,0.0007944709,0.9724473,0.001462731,0.0001134389,0.0004159839,0.0005453652,0.00111473,0.0006297916],"genre_scores_gemma":[0.2913272,0.0007228263,0.7021579,0.0003638683,0.0003285599,0.001035879,0.001759618,0.000400333,0.001903762],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0550755,"threshold_uncertainty_score":0.2912706,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2749071077727216,"score_gpt":0.388502096364703,"score_spread":0.1135949885919814,"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."}}