{"id":"W1994493193","doi":"10.5555/2819009.2819026","title":"Online defect prediction for imbalanced data","year":2015,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":156,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Commit; Process (computing); Machine learning; Reliability (semiconductor); Data mining; Statistical classification; Predictive modelling; Resampling; Software bug; Artificial intelligence; Change detection; Software; Database","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":[0.007483789,0.001107135,0.001262644,0.005291683,0.000661167,0.001361054,0.00142646,0.001223704,0.0007302096],"category_scores_gemma":[0.02960123,0.0003332648,0.0006673695,0.003029001,0.0006228384,0.002338015,0.001186245,0.002215573,0.0007584977],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009294336,"about_ca_system_score_gemma":0.0006172639,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00396743,"about_ca_topic_score_gemma":0.003106352,"domain_scores_codex":[0.9950859,0.001896,0.0003419127,0.001100769,0.001232956,0.0003424346],"domain_scores_gemma":[0.9473374,0.03187877,0.005381492,0.00810714,0.006234489,0.001060652],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001207402,0.001340046,0.3811529,0.0001731689,0.0004359422,0.0005158589,0.0002724134,0.2389631,0.006035556,0.001225377,0.01752052,0.3511578],"study_design_scores_gemma":[0.00001799591,0.00008800172,0.02112061,0.00001484385,0.00001754653,0.00008002095,0.00006924701,0.9735509,0.002175953,0.002248308,0.0006018548,0.0000146965],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.8438423,0.001123894,0.1441721,0.001847162,0.0004217091,0.0001254381,0.00277599,0.004102427,0.001588931],"genre_scores_gemma":[0.96482,0.0001066099,0.03143863,0.0001371306,0.0001843247,0.00005166101,0.002824715,0.00006807889,0.0003688483],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.007483789,"threshold_uncertainty_score":0.03957856,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1116123255897715,"score_gpt":0.3423711398886294,"score_spread":0.2307588142988579,"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."}}