{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/minimax-filter-learning-to-preserve-privacy","title":"Minimax Filter: Learning to Preserve Privacy from Inference Attacks","arxiv_id":"1610.03577","date":"2016-10-12","proceeding":null,"authors":["Jihun Hamm"],"abstract":"Preserving privacy of continuous and/or high-dimensional data such as images,\nvideos and audios, can be challenging with syntactic anonymization methods\nwhich are designed for discrete attributes. Differential privacy, which\nprovides a more formal definition of privacy, has shown more success in\nsanitizing continuous data. However, both syntactic and differential privacy\nare susceptible to inference attacks, i.e., an adversary can accurately infer\nsensitive attributes from sanitized data. The paper proposes a novel\nfilter-based mechanism which preserves privacy of continuous and\nhigh-dimensional attributes against inference attacks. Finding the optimal\nutility-privacy tradeoff is formulated as a min-diff-max optimization problem.\nThe paper provides an ERM-like analysis of the generalization error and also a\npractical algorithm to perform the optimization. In addition, the paper\nproposes an extension that combines minimax filter and differentially-private\nnoisy mechanism. Advantages of the method over purely noisy mechanisms is\nexplained and demonstrated with examples. Experiments with several real-world\ntasks including facial expression classification, speech emotion\nclassification, and activity classification from motion, show that the minimax\nfilter can simultaneously achieve similar or better target task accuracy and\nlower inference accuracy, often significantly lower than previous methods.","url_abs":"http://arxiv.org/abs/1610.03577v3","url_pdf":"http://arxiv.org/pdf/1610.03577v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"minimax-filter-learning-to-preserve-privacy","repo_url":"https://github.com/jihunhamm/MinimaxFilter","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"emotion-classification","task_name":"Emotion Classification"},{"task_slug":"classification","task_name":"General Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1610.03577","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1610.03577"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jihunhamm/MinimaxFilter","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"unverified":4},"by_repo_kind":{"official":{"samples":4,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"12705aea29776386","entry":"KL_divergence","repo":"jihunhamm/MinimaxFilter","repo_kind":"official","path":"src/sparse_autoencoder.py","file_url":"https://github.com/jihunhamm/MinimaxFilter/blob/HEAD/src/sparse_autoencoder.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"12705aea29776386"}},{"code_sha256_prefix":"f2f1e28b4e07fdc3","entry":"run","repo":"jihunhamm/MinimaxFilter","repo_kind":"official","path":"src/privacyPLS.py","file_url":"https://github.com/jihunhamm/MinimaxFilter/blob/HEAD/src/privacyPLS.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"f2f1e28b4e07fdc3"}},{"code_sha256_prefix":"bc3b410c718adea7","entry":"sigmoid","repo":"jihunhamm/MinimaxFilter","repo_kind":"official","path":"src/sparse_autoencoder.py","file_url":"https://github.com/jihunhamm/MinimaxFilter/blob/HEAD/src/sparse_autoencoder.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bc3b410c718adea7"}},{"code_sha256_prefix":"88d624508c794398","entry":"sigmoid_prime","repo":"jihunhamm/MinimaxFilter","repo_kind":"official","path":"src/sparse_autoencoder.py","file_url":"https://github.com/jihunhamm/MinimaxFilter/blob/HEAD/src/sparse_autoencoder.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"88d624508c794398"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}