{"id":"W3180119292","doi":"10.1007/s10664-021-09982-4","title":"Can Offline Testing of Deep Neural Networks Replace Their Online Testing?: A Case Study of Automated Driving Systems.","year":2021,"lang":"en","type":"article","venue":"PubMed","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":44,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"H2020 European Research Council; Canada Research Chairs; Ministry of Education; National Research Foundation of Korea; Fonds National de la Recherche Luxembourg; National Research Foundation; Natural Sciences and Engineering Research Council of Canada; European Commission","keywords":"Computer science; Test strategy; Online and offline; Manual testing; Orthogonal array testing; Context (archaeology); White-box testing; Model-based testing; Non-regression testing; Machine learning; Software performance testing; Black-box testing; Artificial intelligence; Test case; Operating system; Software","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":[],"consensus_categories":[],"category_scores_codex":[0.001102178,0.0002504858,0.0005546933,0.0001434218,0.0001735882,0.00009584343,0.0006929111,0.00008189635,8.603353e-7],"category_scores_gemma":[0.005749102,0.0002336629,0.00005701861,0.002025386,0.00005240093,0.0002292913,0.0009005017,0.0004158042,9.044518e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007922206,"about_ca_system_score_gemma":0.00007615828,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001886389,"about_ca_topic_score_gemma":0.0007844372,"domain_scores_codex":[0.9970483,0.0006072245,0.0008443379,0.0005977499,0.0003623221,0.0005401276],"domain_scores_gemma":[0.9956144,0.001893906,0.000762232,0.000938847,0.0006420035,0.0001486094],"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":[0.000004708447,0.0003361306,0.1012409,0.00006184822,0.00005753665,0.001794901,0.001014872,0.8306893,0.00004371156,0.00002555559,0.000005575072,0.06472494],"study_design_scores_gemma":[0.0006614766,0.00008591745,0.06112935,0.00004690472,0.00003534821,0.001905509,0.001722186,0.9341823,0.00003571399,0.000006218885,0.00000191752,0.0001871864],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.924444,0.0001674912,0.07330123,0.00009003917,0.0005803338,0.0006908852,0.000003451964,0.0006388158,0.00008378979],"genre_scores_gemma":[0.9805061,3.496987e-7,0.01914625,0.00002587874,0.0001576513,0.0001044927,0.000004603623,0.00003044306,0.00002423713],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.103493,"threshold_uncertainty_score":0.9528494,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03895238907324856,"score_gpt":0.2590227423652258,"score_spread":0.2200703532919773,"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."}}