{"id":"W4386214387","doi":"10.1109/csr57506.2023.10224960","title":"Mitigating Membership Inference Attacks in Machine Learning as a Service","year":2023,"lang":"en","type":"article","venue":"","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec en Outaouais","funders":"","keywords":"Computer science; Inference; Exploit; Machine learning; Artificial intelligence; Computer security; Differential privacy; Classifier (UML); Private information retrieval; Confidentiality; Information sensitivity; Data mining","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":[0.01646601,0.001148184,0.001945807,0.001413793,0.002115285,0.004221874,0.003710257,0.003907877,0.001409923],"category_scores_gemma":[0.04323744,0.0006950695,0.001377673,0.001609976,0.003465396,0.01146839,0.008836339,0.007231233,0.0006413559],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002770429,"about_ca_system_score_gemma":0.003400813,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001229436,"about_ca_topic_score_gemma":0.0007781635,"domain_scores_codex":[0.9759988,0.01015998,0.001053769,0.002139212,0.008666446,0.001981865],"domain_scores_gemma":[0.9608001,0.01751324,0.003897252,0.01408395,0.002469846,0.001235658],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001550999,0.0009175511,0.008520462,0.0003419389,0.0003046628,0.0007674326,0.001796447,0.2855943,0.02112809,0.4341737,0.008066176,0.2368384],"study_design_scores_gemma":[0.00003975485,0.0001484282,0.0003607821,0.00003160288,0.00002497973,0.0002745207,0.0001314057,0.8919247,0.007666669,0.09609285,0.003271599,0.00003270716],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05520217,0.0005776696,0.9366688,0.002502627,0.0001022468,0.0002140674,0.00006349238,0.001916535,0.002752311],"genre_scores_gemma":[0.9261777,0.0002107226,0.0714983,0.0004893898,0.0001095969,0.0001480886,0.00008270307,0.00007519579,0.001208345],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01646601,"threshold_uncertainty_score":0.08708161,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06154676532717464,"score_gpt":0.3232635787481941,"score_spread":0.2617168134210195,"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."}}