{"id":"W7117763198","doi":"10.1109/ic366947.2025.11290529","title":"Hybrid Feature Engineering and Tree-Based Ensembles for Predicting Epidemic Outbreaks","year":2025,"lang":"","type":"article","venue":"","topic":"COVID-19 epidemiological studies","field":"Mathematics","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Feature engineering; Autoregressive integrated moving average; Random forest; Feature (linguistics); Boosting (machine learning); Gradient boosting; Regression; Pipeline (software); Support vector machine; Regression analysis","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":[0.001166368,0.0008965151,0.001146196,0.001255649,0.0003828963,0.0006683809,0.0009582712,0.0006257729,0.001180231],"category_scores_gemma":[0.003141913,0.0003456819,0.0009352949,0.001034071,0.0001493924,0.001180338,0.0006203271,0.001038008,0.0006648949],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004418388,"about_ca_system_score_gemma":0.0006552562,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007589331,"about_ca_topic_score_gemma":0.009204859,"domain_scores_codex":[0.9995938,0.00008730758,0.00003320845,0.0001131738,0.0001084072,0.00006403051],"domain_scores_gemma":[0.9989735,0.0003863505,0.00009037443,0.0001422016,0.0003619282,0.00004565401],"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":[0.0002546567,0.000230165,0.009910567,0.0000513126,0.0001626375,0.0001322763,0.00006643675,0.6176097,0.006275609,0.001471568,0.004149178,0.3596859],"study_design_scores_gemma":[0.000003086645,0.00002530119,0.0005555482,0.000002774413,0.000009328179,0.000013624,0.000004439109,0.9978883,0.0006804944,0.0005707137,0.0002423674,0.000004053513],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1419781,0.0006984813,0.8495703,0.0002906293,0.0001620542,0.00007705185,0.000874093,0.004962598,0.001386611],"genre_scores_gemma":[0.8165654,0.0003028155,0.1790394,0.0001174016,0.00009125425,0.000101456,0.001823547,0.0001341275,0.001824673],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007589331,"threshold_uncertainty_score":0.01509035,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08107798944744515,"score_gpt":0.359115530577185,"score_spread":0.2780375411297399,"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."}}