{"id":"W4390102611","doi":"10.1016/j.dsp.2023.104353","title":"Killing two birds with one stone: Quantization achieves privacy in distributed learning","year":2023,"lang":"en","type":"article","venue":"Digital Signal Processing","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Victoria","funders":"Research Grants Council, University Grants Committee; National Natural Science Foundation of China","keywords":"Computer science; Differential privacy; Distributed learning; Quantization (signal processing); Stochastic gradient descent; Federated learning; Artificial intelligence; Distributed computing; Theoretical computer science; Algorithm","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","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0003579578,0.0002090026,0.0002194376,0.0003526184,0.0002420241,0.001216957,0.006211962,0.00008160797,0.000002390038],"category_scores_gemma":[0.002879404,0.000193849,0.00002703082,0.003229188,0.0001137024,0.003691988,0.01307603,0.0004247563,0.0000594557],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009533848,"about_ca_system_score_gemma":0.0001195681,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000139238,"about_ca_topic_score_gemma":0.000003804306,"domain_scores_codex":[0.9979996,0.0000375438,0.0003195349,0.0006155815,0.0004979798,0.0005297231],"domain_scores_gemma":[0.9983607,0.0001502031,0.0001859361,0.001140774,0.00009888844,0.00006350767],"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":[0.00004968697,0.000190886,0.1069155,0.0001951181,0.00002910513,0.0001507173,0.0005402789,0.01581487,0.003080453,0.0009697698,0.0005096347,0.871554],"study_design_scores_gemma":[0.000430887,0.00009230658,0.005394132,0.0004431529,0.000003618507,0.00001144933,0.0001511764,0.9667395,0.001570783,0.02459083,0.0002355014,0.000336664],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2175456,0.00008050827,0.7778665,0.001964385,0.00002565772,0.0001246437,0.00001043652,0.002142414,0.0002398378],"genre_scores_gemma":[0.9831265,0.00001016071,0.01657978,0.00002025898,0.00002640507,0.00001558254,0.0001701767,0.00002896686,0.00002221195],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9509246,"threshold_uncertainty_score":0.9998199,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03206867446080338,"score_gpt":0.2791864865554489,"score_spread":0.2471178120946455,"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."}}