Privacy Preserving Data Mining
By Jaideep Vaidya, Yu Zhu, Chris Clifton. 2015
3,344 citations recorded by the source corpus.
Recent interest in data collection and monitoring using data mining for security and business-related applications has raised privacy. Privacy Preserving Data Mining (PPDM) techniques require data modification to disinfect them from sensitive information or to anonymize them at an uncertainty level. This study uses PPDM with adult dataset to investigate effects of K-anonymization for evaluation metrics. This study uses Artificial Bee Colony (ABC) algorithm for feature generalization and suppression where features are removed without affecting classification accuracy. Also k-anonymity is accomplished by original dataset generalization.
Subfields: Artificial Intelligence, Privacy & Security
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Philip S. Yu
- Vitaly Shmatikov
- Reza Shokri
- Qiang Yang
- Dawn Song
- Peter Kairouz
- Charu C. Aggarwal
- Abhradeep Thakurta
In Privacy & Security
- Peter Kairouz
- Kamalika Chaudhuri
- Adam D. Smith
- Abhradeep Thakurta
- Cynthia Dwork
- Vitaly Shmatikov
- Daniel Kifer
- Kunal Talwar
Citations in the corpus
- Advances and Open Problems in Federated Learning (2019, methodology): “Cryptographic methods for computing on encrypted data were developed starting in the early 1980s [2], [3], and Agrawal and Srikant [4] and Vaidya et al.”
- Slalom: Fast, Verifiable and Private Execution of Neural Networks in Trusted Hardware (2018, background): “, in the latter two where the computation is performed locally by the data owner), the need for integrity is a common denominator, and this is the property we primarily focus on in this paper (and for which we obtain the best results!) Cryptography can address some of these problems [17, 32, 3, 38, ”
- Privacy-Preserving Distributed Linear Regression on High-Dimensional Data (2017, background): “Questions of privacy-preserving data mining and private computation of machine learning algorithms have been considered in several works [19, 40, 48, 71], providing theoretical protocols without implementations or practical evaluations of their efficiency.”
- Distributed Statistical Machine Learning in Adversarial Settings (2017, background): “Although a variety of privacypreserving solutions have been developed [4, 11], privacy breaches still occur frequently, with recent examples including iCloud leaks of celebrity photos [3] and PRISM surveillance program [2].”
- Federated Learning: Challenges, Methods, and Future Directions (2019, background): “Privacy-preserving learning has been extensively studied by the machine learning [e.g., 76], systems [e.g., 4, 11], and theory [e.g., 38, 69] communities.”
- Machine Learning Models that Remember Too Much (2017, background): “Startingwith [49], there has beenmuch research on using secure multi-party computation to enable several parties to create a joint model on their separate datasets, e.”
- Practical Evasion of a Learning-Based Classifier: A Case Study (2014, methodology): “I. INTRODUCTION Data analysis methods such as machine learning are increasingly used in security applications.”
- Hacking smart machines with smarter ones: How to extract meaningful data from machine learning classifiers (2013, background): “Differential privacy [19, 14, 9] protects against unintentional disclosure of potentially sensitive information related to a single record of a database D .”
- The Value of Collaboration in Convex Machine Learning with Differential Privacy (2019, methodology): “In the past, secure multi-party computation has been used in various ML models, such as decision trees [5], regression [6], association rules [7], and clustering [8],”
- On the Difficulties of Disclosure Prevention in Statistical Databases or The Case for Differential Privacy (2010, positive): “The work of Agrawal and Srikant [2] and the spectacular privacy compromises achieved by Sweeney [33] rekindled interest in the problem among computer scientists, particularly within the database community.”
- Auditing black-box models for indirect influence (2016, background): “There are potential connections to privacy-preserving data mining [1]: however, the trust model is different: in privacy-preserving data mining, we do not trust the user of the results of classification with sensitive information from the input.”
- Data Synthesis based on Generative Adversarial Networks (2018, methodology): “Moreover, methods have been developed to recognize and remove noise [10].”
- LOGAN: Membership Inference Attacks Against Generative Models (2017, methodology): “Privacy-enhancing tools based on secure multiparty computation and homomorphic encryption have been proposed to securely train supervised machine learning models, such as decision trees [36], linear regressors [15], and neural networks [9, 14].”
- Machine Learning with Membership Privacy using Adversarial Regularization (2018, methodology): “ganduseofmachinelearningmodels. These methods leverage various techniques including homomorphicencryption, garbledcircuits,and securemulti-partycomputationfor privatemachine learning onencrypted data [8, 20, 33, 37], as well as private computation using trusted hardware (e.g., Intel SGX)[26,39].Alth”
- Composition attacks and auxiliary information in data privacy (2008, neutral): “Randomization-based schemes introduce uncertainty either by randomly perturbing the raw data (a technique called input perturbation , randomized response , e.g., [34, 2, 16]), or post-randomization , e.g., [32]), or by injecting randomness into the algorithm used to analyze the data (e.g., [6, 28]).”
- LOGAN: Evaluating Privacy Leakage of Generative Models Using Generative Adversarial Networks (2017, methodology): “Defense mechanisms based on secure multiparty computation and homomorphic encryption have been proposed to train supervised machine learning models, such as decision trees [26], linear regressors [11], and neural networks [6, 10].”
- Privacy-Preserving Multivariate Statistical Analysis: Linear Regression and Classification (2004, background): “Although some work has extended SMC to more complicated computation problems, such as data mining problem in [18, 25], most of them are still not efficient enough for practical use.”