{"id":"W4406502827","doi":"10.1093/imaiai/iaae034","title":"Differentially private low-dimensional synthetic data from high-dimensional datasets","year":2025,"lang":"en","type":"article","venue":"Information and Inference A Journal of the IMA","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Curse of dimensionality; Principal component analysis; Synthetic data; Intrinsic dimension; Covariance; Computer science; Clustering high-dimensional data; Key (lock); Dimension (graph theory); Algorithm; Covariance matrix; Data mining; Artificial intelligence; Mathematics; Statistics","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":[0.007216215,0.0006752547,0.001119227,0.000829206,0.0009810199,0.002177061,0.001800571,0.001596008,0.001958527],"category_scores_gemma":[0.02962595,0.0005508077,0.0009011711,0.001327046,0.002779382,0.003945714,0.004643739,0.002984357,0.0006160769],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001425373,"about_ca_system_score_gemma":0.001318465,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004099323,"about_ca_topic_score_gemma":0.0003394309,"domain_scores_codex":[0.9939039,0.003152244,0.0002616051,0.0007350387,0.001642078,0.0003050504],"domain_scores_gemma":[0.9812918,0.009428673,0.001325962,0.006464489,0.00108473,0.0004043325],"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":[0.001109948,0.000203892,0.003945482,0.0002121743,0.0001312308,0.0004131576,0.0003711685,0.5389113,0.01552592,0.3522384,0.005504041,0.08143319],"study_design_scores_gemma":[0.00003649649,0.00007454228,0.0003865287,0.00002011288,0.000011247,0.0001922802,0.00005018474,0.8353574,0.007642143,0.1542941,0.001913589,0.00002142298],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03337327,0.0001404848,0.9630647,0.001092236,0.00005923837,0.00009248091,0.000426814,0.0003125487,0.00143833],"genre_scores_gemma":[0.8102618,0.0002541174,0.1853334,0.0004680023,0.0001459402,0.0002459625,0.001182787,0.00008542542,0.002022621],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007216215,"threshold_uncertainty_score":0.03816342,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02040655346756855,"score_gpt":0.272225933458451,"score_spread":0.2518193799908824,"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."}}