Secure methods learn fair models without revealing sensitive attributes.
problem Training fair machine learning models without exposing sensitive data.
method Secure multi-party computation to encrypt sensitive attributes.
result Outcome-based fair models can be learned, checked, or verified without revealing sensitive attributes.
A framework for partially encrypted machine learning using functional encryption.
problem Performing machine learning on encrypted data without revealing sensitive information.
method Combining adversarial training and functional encryption to efficiently compute quadratic functions and prevent feature leakage.
result The proposed framework maintains high model accuracy while significantly improving data privacy.
Securely trains fair models using homomorphic encryption.
problem Protecting sensitive features while testing model fairness.
method Fully homomorphic encryption for training and testing.
result Practical application to adult income data set.
Improved CNN accuracy for encrypted data using approximate activation functions.
problem Low accuracy in classifying encrypted data using homomorphic encryption.
method Used a fourth-order polynomial approximation of the Swish activation function with batch normalization for homomorphic encryption.
result Achieved 99.22% accuracy on MNIST and 80.48% on CIFAR-10, improving by 0.04% and 4.11% respectively.
Paper proposes efficient privacy-preserving matrix encryption for secure collaborative learning against malicious adversaries.
problem Secure collaborative learning of sensitive data across different agencies is challenging with malicious adversaries.
method The paper applies matrix encryption to secure data against chosen plaintext attack, known plaintext attack, and collusion attack, achieving local differential privacy and high computation efficiency.
result The proposed schemes are computationally efficient and secure against malicious adversaries compared to existing techniques.
Paper proposes DPN for encrypted speech recognition, maintaining privacy and security.
problem Privacy and security issues in cloud-based speech recognition.
method Deep Polynomial Network (DPN) for encrypted speech, cloud-local joint decoding.
result DPN can make frame-level predictions over encrypted speech and return them in encrypted form.
Securely analyzes survival data across multiple institutions without revealing individual patient records.
problem Privacy concerns in federated survival analysis of health data.
method Multiparty homomorphic encryption for approximate floating-point computation and encrypted aggregation.
result Privacy-preserving federated Kaplan--Meier survival analysis with high fidelity and predictable overhead.
Paper proposes privacy-preserving learning for images, making them imperceptible to humans but recognizable by machines.
problem Conflict between developing AI systems and protecting sensitive training data.
method Encryption strategies (random shuffling and sub-patch mixing) followed by minimal adaptation to vision transformer.
result Achieves comparable accuracy to competitive methods while ensuring human-imperceptibility of encrypted images.
RENNs protect input privacy by rotating d-ary features.
problem Protecting input privacy from intermediate-layer features.
method Rotation-equivariant neural networks using d-ary vectors/tensors.
result RENNs effectively hide input information without degrading output accuracy.
Modular FHE enables accurate Gaussian process predictions without data exposure.
problem Privacy issues in machine learning with multiple data sources.
method Modular approach to apply FHE only to sensitive parts of a workflow.
result First effectively encrypted Gaussian process model.
TG-PSM morphs encrypted traffic to reduce traffic analysis accuracy.
problem Traffic analysis on encrypted traffic reveals sensitive information.
method Clusters websites, finds target sites, morphs traffic using greedy algorithm.
result TG-PSM reduces attacker's accuracy significantly with low overhead.
Paper develops methods for fair insurance pricing without direct access to sensitive attributes.
problem Fairness in insurance pricing with restricted access to sensitive attributes.
method Develops statistical methods for estimating discrimination-free premiums using privatized sensitive attributes.
result The proposed methods enable fair insurance pricing while respecting privacy and regulatory constraints.
A fair clustering method for multiple sensitive attributes is proposed.
problem Ensuring fair representation of sensitive attributes in clustering.
method FairKM (Fair K-Means) method inspired by K-Means, using fairness and coherence objectives.
result FairKM clusters significantly better on both quality and fair representation of sensitive attribute groups.
Paper introduces privacy-preserving deep-learning service.
problem Privacy loss in deep-learning-as-a-service.
method Homomorphic Encryption tailored for CNNs.
result Effectiveness of proposed privacy-preserving architecture.
TaCo prevents non-linear classifiers from detecting sensitive attributes.
problem Ensuring fairness in NLP models by preventing sensitive attribute detection.
method Targeted Concept Erasure (TaCo) removes sensitive information from final latent representations, even against non-linear classifiers.
result TaCo outperforms state-of-the-art methods in reducing sensitive attribute prediction accuracy while preserving overall task performance.
Topological data analysis classifies encrypted bits with success.
problem Classifying encrypted data with traditional machine learning methods.
method Persistent homology for generating topological features, machine learning pipeline.
result Successfully classifies encrypted data, outperforming classical models.
Information that is stored in an encrypted format is, by definition, usually not amenable to statistical analysis or machine learning methods. In this paper we present detailed analysis of coordinate and accelerated gradient descent algorithms which are capable of fitting least squares and penalised ridge regression mo…
This work introduces a fair learning method for diverse sensitive attributes.
problem Fairness in supervised learning with complex sensitive attributes.
method Neural network with a simple random sampler for fairness penalties.
result The method improves fairness and utility on benchmark data.
TAPAS speeds up encrypted machine learning predictions.
problem Data privacy and computation time in encrypted machine learning.
method Combining binarization, sparsification, and homomorphic encryption techniques.
result Significantly reduces computation time for encrypted predictions.
Paper proposes a new FRL algorithm for continuous sensitive attributes using EIPM.
problem Existing FRL algorithms cannot handle continuous sensitive attributes.
method Introduces EIPM to assess fairness in representation space for continuous attributes and proposes FREM algorithm.
result FREM outperforms other methods in fairness evaluation for continuous sensitive attributes.
Framework for neural network inference on encrypted data.
problem Privacy-preserving machine learning with homomorphic encryption.
method Extensible framework using TensorFlow and SEAL, exposing algebra and deep learning primitives.
result Framework allows training and inference of neural networks on encrypted data.
FairICP addresses equalized odds fairness for multiple sensitive attributes.
problem Equalized odds fairness for multiple sensitive attributes.
method Adversarial learning with inverse conditional permutation.
result Promotes equalized odds under complex, multi-dimensional sensitive attributes.
Proposes a sequential framework for fairness in multiple sensitive attributes.
problem Fairness in the presence of multiple sensitive attributes.
method Sequential framework using multi-marginal Wasserstein barycenters.
result Closed-form solution for sequentially fair predictor.
The paper shows how to use proxy attributes for fairness in machine learning models with missing sensitive group data.
problem Measuring and enforcing fairness in machine learning models with incomplete sensitive group data.
method Using proxy-sensitive attributes to derive upper bounds on multiaccuracy and multicalibration violations and adjust models to satisfy these fairness notions.
result Provable upper bounds on multiaccuracy and multicalibration violations can be derived using proxy-sensitive attributes in the absence of sensitive group data.
Scalable learning of nodal attributes in large graphs with privacy.
problem Scalability and privacy in learning nodal attributes of large graphs.
method Multikernel-based approach for real-time evaluation of nodal attributes.
result Real-time evaluation of nodal attributes without re-solving the problem over all nodes.
Study on protecting sensitive properties of datasets during analysis.
problem Ensuring privacy of sensitive properties in datasets.
method Proposes definitions and mechanisms for attribute privacy using the Pufferfish framework.
result Developed efficient and inefficient mechanisms for attribute privacy.
A method for data encryption makes data look identical to humans but misleading to machine learning.
problem Data leakage in data sharing for medical purposes.
method Proposes a method inspired by adversarial attacks for data encryption.
result Encrypted data look identical to humans but misleading to machine learning methods.
Introduces FairCOCCO for fair learning with multitype, multivariate sensitive attributes.
problem Fairness in machine learning with multiple, complex sensitive attributes.
method FairCOCCO measure based on cross-covariance operators, incorporating a regularisation term.
result Consistent improvements in balancing fairness and predictive power on real-world datasets.
New algorithm tackles subgroup fairness in AI with multiple sensitive attributes.
problem Heavy computational burdens and data sparsity in subgroup fairness for multiple sensitive attributes.
method Doubly Regressing Adversarial learning (DRAF) for subgroup fairness, focusing on subgroups with sufficient sample sizes and marginal fairness.
result DRAF algorithm reduces a surrogate fairness gap for supIPM with less computation than directly reducing supIPM.
Deep learning attacks chaos-based image encryption.
problem Chaos-based image encryption vulnerability.
method Project encrypted images to low-dimensional space, use deconvolutional generator to regenerate images.
result Proposes a key-independent, end-to-end trained method to attack chaos-based encryption.
Study shows data attribution methods are sensitive to hyperparameters, making tuning costly.
problem Hyperparameter sensitivity in data attribution methods makes tuning impractical.
method Theoretical analysis and lightweight procedure for selecting regularization value without retraining.
result Proposes a lightweight procedure for selecting regularization value without model retraining.
Paper defends sensitive attributes in GNNs from inference attacks.
problem Protecting sensitive attributes in GNNs from inference attacks.
method Proposes adversarial training with TV and Wasserstein distance to locally filter sensitive attributes.
result Framework creates strong defense against inference attacks with minimal performance loss.
New technique reduces bias in DNN models without sensitive attribute annotations.
problem Existing bias mitigation methods require instance-level annotations and do not guarantee removal of all sensitive information.
method Representation Neutralization for Fairness (RNF) debiases only the classification head of DNN models using neutralized representations.
result RNF effectively reduces discrimination of DNN models with minimal performance degradation.
Cryptotree enables accurate predictions on encrypted data using Random Forests.
problem Applying machine learning to private data while preserving confidentiality.
method Adapting Neural RF to CKKS scheme for HE operations on encrypted data.
result Cryptotree achieves better prediction results on encrypted data than regular RF.
Private release of sensitive data enables fair learning.
problem Learning fair predictors with restricted sensitive data.
method Private release of sensitive demographic data, adapting non-discriminatory learners.
result The approach provides theoretical guarantees on performance for fair predictors.
Overlearning exposes hidden, sensitive attributes in models, threatening privacy and bias.
problem Models learn unintended, sensitive attributes beyond their training objectives.
method Demonstrated and analyzed overlearning in vision and NLP models.
result Overlearning reveals sensitive attributes that break privacy protections and can be re-purposed for harmful tasks.
New method stabilizes inputs to DNN for secure inference with LHE.
problem Incompatibility of LHE with nonlinear functions in DNN.
method Training with polynomial approximations and Min-Max normalization.
result Loss in prediction accuracy reduced to small values or eliminated.
Glyph speeds up DNN training on encrypted data by 99%.
problem Training deep neural networks on encrypted data efficiently.
method Switches between TFHE and BGV cryptosystems for activations and MAC operations, applies transfer learning.
result Reduces training latency by 99% on various encrypted datasets.
New method optimizes fairness in predictive models for continuous sensitive attributes.
problem Enforcing full statistical independence on continuous sensitive attributes is too restrictive.
method Functional bilevel optimization (FBO) and ITD algorithms.
result Achieves lowest or near-lowest fairness-accuracy regret on synthetic and real datasets.
Develops a method to ensure fairness across multiple sensitive attributes in machine learning.
problem Ensuring fairness among demographic groups formed by multiple sensitive attributes.
method Formulates intersectional fairness as a mutual information minimization problem and proposes a generic end-to-end algorithmic framework.
result Demonstrates effective debiasing of classification results with minimal impact to accuracy.
Algorithmic decision making process now affects many aspects of our lives. Standard tools for machine learning, such as classification and regression, are subject to the bias in data, and thus direct application of such off-the-shelf tools could lead to a specific group being unfairly discriminated. Removing sensitive …
Local explanation methods for deep networks are found to be insensitive to parameter values.
problem Local explanation methods lack sensitivity to parameter values in deep neural networks.
method Investigated the sensitivity of local explanation methods (e.g., integrated gradients) to parameter values in randomly-initialized and learned networks.
result IG attributions for a random network and the actual network are uncorrelated when both factors (signs and baseline pixels) are accounted for.
Recent advances in cryptography promise to enable secure statistical computation on encrypted data, whereby a limited set of operations can be carried out without the need to first decrypt. We review these homomorphic encryption schemes in a manner accessible to statisticians and machine learners, focusing on pertinent…
Develops a framework to obfuscate sensitive attributes in machine learning models.
problem Minimizing information leakage of sensitive attributes in crowdsourced data.
method Proposes a minimax optimization formulation and proves an information-theoretic lower bound.
result Adversarial learning achieves the best trade-off between attribute obfuscation and accuracy.
EXOC framework uses auxiliary variables for counterfactual fairness in machine learning.
problem Balancing fairness and predictive accuracy in models with sensitive attributes.
method EXOC framework uses auxiliary variables to define an auxiliary node and a control node for counterfactual fairness.
result EXOC framework outperforms state-of-the-art approaches in achieving counterfactual fairness.
Two solutions improve privacy-preserving inference with reduced latency and wider neural network support.
problem Balancing accuracy, security, and computational complexity in machine learning with sensitive data.
method Combining Homomorphic Encryption with transfer learning and novel data representation methods.
result More than 10x improvement in latency with wider neural network support.
New findings show local attributions can't be both robust and provide recourse.
problem Ensuring machine learning systems are accountable and provide actionable recourse options.
method Formal definition of recourse sensitivity and counterexamples for popular attribution methods.
result It is impossible for any single attribution method to be both robust and provide recourse.
End-to-end encrypted neural network improves privacy and compression in federated learning.
problem Privacy and bandwidth issues in gradient updates transmission in federated learning.
method Proposes an end-to-end encrypted neural network to encode and decode gradient updates.
result Effective privacy protection and data compression with minimal accuracy loss.