{"id":"W4402390737","doi":"10.23889/ijpds.v9i5.2518","title":"Leveraging Machine Learning to Combat Missingness and Error in Data","year":2024,"lang":"en","type":"article","venue":"International Journal for Population Data Science","topic":"Anomaly Detection Techniques and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Victoria Park","funders":"","keywords":"Missing data; Computer science; Artificial intelligence; Machine learning","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":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.001494039,0.00007257015,0.00006842954,0.0004941182,0.0003228987,0.00154678,0.003404221,0.00001874124,0.000006679077],"category_scores_gemma":[0.0002687286,0.00006696417,0.00001212328,0.000591529,0.00004110516,0.003946093,0.001556845,0.0001751432,0.000004640075],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009151742,"about_ca_system_score_gemma":0.00009586851,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001644522,"about_ca_topic_score_gemma":0.00004878252,"domain_scores_codex":[0.9986163,0.00001813169,0.0002661659,0.0005299501,0.000417693,0.0001517493],"domain_scores_gemma":[0.9991292,0.00007415366,0.00006987332,0.0005036864,0.0001251466,0.00009787617],"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":[0.00001873643,0.00006321211,0.01177239,0.00002109156,0.00002140334,0.00003265316,0.0009086651,0.003216078,0.004485236,0.06700683,0.002832877,0.9096208],"study_design_scores_gemma":[0.0000786767,0.00001435769,0.008179114,0.00006392325,0.00000208113,0.0001599,0.00002718943,0.9240571,0.00008999627,0.003323105,0.06391559,0.00008902024],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02735101,0.0001499498,0.9639165,0.007284448,0.000944985,0.0001368899,0.00006901108,0.0001004426,0.00004681022],"genre_scores_gemma":[0.9055858,0.00003143297,0.09381937,0.0001722111,0.0001177717,0.000005908751,0.0001154457,0.000005745294,0.0001463513],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.920841,"threshold_uncertainty_score":0.9994897,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1132270631980558,"score_gpt":0.4198903725605341,"score_spread":0.3066633093624783,"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."}}