A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
CheckNet verifies neural network inference on untrusted devices.
Federated learning leaks participant dataset quality even with secure aggregation.
Secure neural network inference on untrusted platforms using holographic reduced representations.
Optimizes crypto-oriented neural architectures for faster secure inference.
The exponential increase in dependencies between the cyber and physical world leads to an enormous amount of data which must be efficiently processed and stored. Therefore, computing paradigms are evolving towards machine learning (ML)-based systems because of their ability to efficiently and accurately process the eno…
This research highlights the secrecy potential of nonlinear generative models and their all-or-nothing phase transition.
This paper optimizes SMPC for neural network inference, reducing memory and time.
New method uses model's generalization gap to predict membership inference attacks.
SLIP secures LLMs on edge devices by splitting computation and protecting sensitive parts.
This article reviews recent advances in secure distributed and decentralized inference and learning against Byzantine threats.
We create a formal framework for the design of informative securities in prediction markets. These securities allow a market organizer to infer the likelihood of events of interest as well as if he knew all of the traders' private signals. We consider the design of markets that are always informative, markets that are …
In machine learning (ML) security, attacks like evasion, model stealing or membership inference are generally studied in individually. Previous work has also shown a relationship between some attacks and decision function curvature of the targeted model. Consequently, we study an ML model allowing direct control over t…
This paper improves federated learning efficiency by auto-tuning secure aggregation parameters.
Bayesian methods improve adversarial machine learning robustness.
The arms race between attacks and defenses for machine learning models has come to a forefront in recent years, in both the security community and the privacy community. However, one big limitation of previous research is that the security domain and the privacy domain have typically been considered separately. It is t…
Paper defends diffusion models from membership inference attacks using Langevin dynamics.
When applying machine learning to sensitive data, one has to find a balance between accuracy, information security, and computational-complexity. Recent studies combined Homomorphic Encryption with neural networks to make inferences while protecting against information leakage. However, these methods are limited by the…
Study quantized models' privacy against membership inference attacks.
Broker uses multi-task dynamic pricing to learn competitive prices in credit markets.
Security, privacy, and fairness have become critical in the era of data science and machine learning. More and more we see that achieving universally secure, private, and fair systems is practically impossible. We have seen for example how generative adversarial networks can be used to learn about the expected private …
The emerging paradigm of Human-Machine Inference Networks (HuMaINs) combines complementary cognitive strengths of humans and machines in an intelligent manner to tackle various inference tasks and achieves higher performance than either humans or machines by themselves. While inference performance optimization techniqu…
SecureGBM securely trains GBM models across two parties without revealing data.
New research limits how well attackers can guess if data points were in a model's training set.
The paper addresses challenges in edge deep learning for IoT, proposing new directions.
Paper introduces attacks to infer GAN training dataset properties.
Defends against ML inference attacks using adversarial examples.
To promote secure and private artificial intelligence (SPAI), we review studies on the model security and data privacy of DNNs. Model security allows system to behave as intended without being affected by malicious external influences that can compromise its integrity and efficiency. Security attacks can be divided bas…
Multilayer networks are attracting growing attention in many fields, including finance. In this paper, we develop a new tractable procedure for multilayer aggregation based on statistical validation, which we apply to investor networks. Moreover, we propose two other improvements to their analysis: transaction bootstra…
Most of the data manipulation attacks on deep neural networks (DNNs) during the training stage introduce a perceptible noise that can be catered by preprocessing during inference or can be identified during the validation phase. Therefore, data poisoning attacks during inference (e.g., adversarial attacks) are becoming…
Recurrent neural networks and sequence to sequence models require a predetermined length for prediction output length. Our model addresses this by allowing the network to predict a variable length output in inference. A new loss function with a tailored gradient computation is developed that trades off prediction accur…
SOTERIA optimizes neural networks for secure inference with minimal overhead.
Improved fraud detection in finance with quantum-enhanced federated learning.
DeepPeep attacks DNN architectures to reveal design details, posing IP theft risks.
Leveled Homomorphic Encryption (LHE) offers a potential solution that could allow sectors with sensitive data to utilize the cloud and securely deploy their models for remote inference with Deep Neural Networks (DNN). However, this application faces several obstacles due to the limitations of LHE. One of the main probl…
Federated learning facilitates the collaborative training of models without the sharing of raw data. However, recent attacks demonstrate that simply maintaining data locality during training processes does not provide sufficient privacy guarantees. Rather, we need a federated learning system capable of preventing infer…
As Machine Learning (ML) gets applied to security-critical or sensitive domains, there is a growing need for integrity and privacy for outsourced ML computations. A pragmatic solution comes from Trusted Execution Environments (TEEs), which use hardware and software protections to isolate sensitive computations from the…
Optimizes a portfolio for an investor preferring accepted securities over a reference security.
CHEETAH speeds up secure MLaaS by 100x over fastest existing schemes.
Study shows Bitcoin security tied to mining rewards and prices.
A pair of points in a riemannian manifold makes a secure configuration if the totality of geodesics connecting them can be blocked by a finite set. The manifold is secure if every configuration is secure. We investigate the security of compact, locally symmetric spaces.
Cronus securely transfers model parameters to protect federated learning from poisoning attacks.
We say that a pair of points x and y is secure if there exist a finite set of blocking points such that any geodesic between x and y passes through one of the blocking points. The main point of this paper is to exhibit new examples of blocking phenomena both in the manifold and the billiard table setting. As an approac…
Industry lacks tools to secure ML systems, study finds.
Differentially private method for synthetic data generation from vertically partitioned data.
Sparse oblique decision tree improves security rules for renewable power systems.
RL models improve target control in SSGs for security applications.
The paper evaluates methods for explaining deep learning in security.
A riemannian manifold is secure if the geodesics between any pair of points in the manifold can be blocked by a finite number of point obstacles. Compact, flat manifolds are secure. A standing conjecture says that these are the only secure, compact riemannian manifolds. The conjecture claims, in particular, that a riem…