{"id":"W4410614229","doi":"10.1109/tdsc.2025.3572527","title":"The Power of Bias: Optimizing Client Selection in Federated Learning With Heterogeneous Differential Privacy","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Dependable and Secure Computing","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"Basic and Applied Basic Research Foundation of Guangdong Province; National Natural Science Foundation of China","keywords":"Computer science; Differential privacy; Selection (genetic algorithm); Information privacy; Selection bias; Computer network; Computer security; Data mining; Artificial intelligence","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":[],"consensus_categories":[],"category_scores_codex":[0.0003205625,0.0001778204,0.0002126979,0.0002514416,0.000726618,0.0002712722,0.002175,0.0001051267,0.000002796211],"category_scores_gemma":[0.000205289,0.0001336967,0.00004408706,0.0007996642,0.00006938551,0.0002079302,0.0004044639,0.0006013925,7.423528e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006936748,"about_ca_system_score_gemma":0.00006812983,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006509304,"about_ca_topic_score_gemma":0.0001273195,"domain_scores_codex":[0.9985422,0.0001528927,0.0003308956,0.0004363065,0.0002031142,0.0003345296],"domain_scores_gemma":[0.9986035,0.0004123983,0.0001294579,0.0007451024,0.0000787637,0.00003082971],"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":[0.0003997697,0.0006182517,0.002289222,0.000180752,0.0004386711,0.00006370793,0.001872891,0.6500605,0.008684087,0.0009473251,0.0004833425,0.3339615],"study_design_scores_gemma":[0.0005702542,0.0002146351,0.0002014294,0.0002252013,0.00001239584,0.00003738524,0.0001451439,0.9498695,0.04777637,0.0006715989,0.0001178848,0.000158171],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3603168,0.0000768585,0.6385731,0.0005227808,0.0001559713,0.0001320236,9.38446e-7,0.0001724154,0.00004917676],"genre_scores_gemma":[0.9767862,0.00006666893,0.02307631,0.0000256762,0.0000048426,0.000007445128,7.28531e-7,0.000009683737,0.00002246076],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6164694,"threshold_uncertainty_score":0.5588632,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01580447681024537,"score_gpt":0.2497590021081102,"score_spread":0.2339545252978648,"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."}}