{"id":"W4404295861","doi":"10.36227/techrxiv.173143474.47669751/v1","title":"Exploring the Preprocessing of a Time-Series Administrative Healthcare Dataset on Deep Learning to Improve Prediction","year":2024,"lang":"en","type":"preprint","venue":"","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Hotchkiss Brain Institute; University of Calgary","funders":"Canadian Institutes of Health Research; University of Calgary; Alberta Innovates; University of Alberta; Government of Alberta; Alberta Health Services","keywords":"Preprocessor; Computer science; Deep learning; Time series; Series (stratigraphy); Artificial intelligence; Data pre-processing; Health care; Machine learning; Data mining; Data science; Political science","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.001135837,0.0004166783,0.0004469275,0.0002673634,0.0003495135,0.0004676761,0.001583284,0.0001647633,0.00002419687],"category_scores_gemma":[0.0006188701,0.0003086424,0.0001087938,0.0005149513,0.00006709454,0.0003701721,0.003917694,0.002664022,0.0001026867],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001956039,"about_ca_system_score_gemma":0.000572641,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008035776,"about_ca_topic_score_gemma":0.0001464399,"domain_scores_codex":[0.9962192,0.0004750662,0.000722174,0.00143474,0.0007074345,0.000441443],"domain_scores_gemma":[0.9971358,0.0003746606,0.0003844501,0.001655915,0.0002492656,0.0001999065],"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":[0.0004217977,0.0001992701,0.001902349,0.01455131,0.0003774026,0.0001036182,0.10298,0.1551242,0.0005986154,0.0311045,0.004642405,0.6879945],"study_design_scores_gemma":[0.0004646693,0.007204182,0.006637293,0.01004129,0.0001465787,0.0001078041,0.006185217,0.9094114,0.01124684,0.02160926,0.02473675,0.002208739],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2714346,0.00449231,0.4335418,0.2124092,0.0241936,0.01602859,0.007542941,0.009641969,0.02071501],"genre_scores_gemma":[0.9614415,0.0001161103,0.03356796,0.000747669,0.0006887255,0.001104257,0.0007674759,0.00008990279,0.001476369],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7542871,"threshold_uncertainty_score":0.9999366,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09581581284355277,"score_gpt":0.3572450434689112,"score_spread":0.2614292306253584,"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."}}