Research
On-device research index

arXiv research

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.

168,742 papers · 148 categories

Trend · papers per month

101203304405 · Jun 202019922001200920172026
48 results for security vs accuracy

Adaptive Misinformation defends against model stealing attacks by sending incorrect predictions for OOD queries.

problem Model stealing attacks clone target models using black-box query access and a surrogate dataset.
method Selective sending of incorrect predictions for Out-Of-Distribution (OOD) queries to degrade attacker's clone model accuracy.
result Our defense reduces attacker's clone model accuracy by up to 40% while maintaining benign user accuracy under 0.5%.

A system for federated learning with private data, adding discrete Gaussian noise and secure aggregation.

problem Training models on private data distributed across devices while ensuring privacy.
method Discretizes data, adds discrete Gaussian noise, and uses secure aggregation to protect privacy.
result Matches the accuracy of central differential privacy with less than 16 bits of precision per value.

We develop a secure aggregation protocol for federated learning that reduces communication and computation costs.

problem Expensive communication and privacy concerns in federated learning.
method Adapting compression-based federated techniques to additive secret sharing.
result Our protocol achieves high accuracy with low communication costs and is more efficient than prior work.

We propose a new randomized ensemble technique with a provable security guarantee against black-box transfer attacks. Our proof constructs a new security problem for random binary classifiers which is easier to empirically verify and a reduction from the security of this new model to the security of the ensemble classi…

2019-06-07abs ↗pdf ↗

SLIP secures LLMs on edge devices by splitting computation and protecting sensitive parts.

problem Protecting LLMs on edge devices from theft and unauthorized use.
method SLIP uses matrix decomposition to split model computation between secure and vulnerable resources, ensuring zero accuracy degradation and minimal latency.
result SLIP is the first practical, secure hybrid protocol for protecting LLMs on edge devices.

Training of one-vs.-rest SVMs can be parallelized over the number of classes in a straight forward way. Given enough computational resources, one-vs.-rest SVMs can thus be trained on data involving a large number of classes. The same cannot be stated, however, for the so-called all-in-one SVMs, which require solving a …

2016-11-25abs ↗pdf ↗

Deep learning is increasingly used as a building block of security systems. Unfortunately, neural networks are hard to interpret and typically opaque to the practitioner. The machine learning community has started to address this problem by developing methods for explaining the predictions of neural networks. While sev…

2019-06-05abs ↗pdf ↗

CoroNet detects COVID-19 from chest X-rays with high accuracy.

problem Detecting COVID-19 from chest X-rays using limited testing kits.
method Proposes CoroNet, a deep neural network based on Xception architecture trained on a combined dataset of COVID-19 and pneumonia X-rays.
result CoroNet achieved an overall accuracy of 89.6% and precision/recall rates of 93%/98.2% for 4-class cases (COVID vs Pneumonia bacterial vs pneumonia viral vs normal).

This paper studies the trade-off between model accuracy and coverage for diagnosis models used by patients.

problem Balancing accuracy and coverage in diagnosis models for patient use.
method Learned diagnosis models with varying coverage from EHR data.
result A 1% drop in top-3 accuracy for every 10 diseases added to the coverage.

Secure XGB for privacy-preserving machine learning in federated learning.

problem Privacy-preserving machine learning in federated learning with practical gradient tree boosting models.
method Secure multi-party computation, distributed model storage, secure permutation protocols.
result Our XGB models provide competitive accuracy and practical performance.

The study reduces a personality measurement instrument to 10 features with minimal loss of accuracy.

problem Accurately predicting personality types with minimal predictors.
method Reduced a 94-item personality survey to 10 features, using class imbalance correction methods.
result Achieved 73.81% accuracy on unseen data with only 1% loss.

Unified market-based description of returns and variances of trades.

problem Market-based variance of trades and market portfolio.
method Unified market-based approach to describe returns and variances of trades and market portfolio.
result Market-based variance accounts for random volumes of trades and differs from Markowitz's portfolio variance.

Secure federated learning framework resists adversarial users.

problem Resilience against adversarial (Byzantine) users in federated learning.
method Integrated stochastic quantization, verifiable outlier detection, and secure model aggregation.
result First single-server Byzantine-resilient secure aggregation framework (BREA) for secure federated learning.

Secure federated learning reduces privacy risks with differential privacy and secure multiparty computation.

problem Reverse engineering of private client data from federated learning model parameters.
method Combining differential privacy and secure multiparty computation.
result Improved accuracy of shared models without significant privacy loss.

Robust cancer screening model using pre-trained ensembles for biomarkers.

problem Detecting early-stage cancer, especially in hard-to-diagnose cases like pancreatic cancer.
method Meta-trained Hyperfast model for robust classification, combined with ensembling of XGBoost and LightGBM.
result Achieved highest AUC of 0.9929 and robust performance on imbalanced datasets.

RS-NN predicts belief functions for classification, improving accuracy and uncertainty estimation.

problem Improving confidence in machine learning predictions for safety-critical applications.
method Random-Set Neural Network (RS-NN) using random set mathematics.
result RS-NN outperforms state-of-the-art methods in accuracy, uncertainty estimation, and OoD detection.

The rapid digital transformation without security considerations has resulted in the rise of global-scale cyberattacks. The first line of defense against these attacks are Network Intrusion Detection Systems (NIDS). Once deployed, however, these systems work as blackboxes with a high rate of false positives with no mea…

2019-08-26abs ↗pdf ↗

Paper designs optimal ECOCs using IP for robust multiclass classification.

problem Designing robust ECOCs for multiclass classification.
method Integer Programming formulation to minimize codebooks with desirable error-correcting properties, leveraging graph-theoretic structure and edge clique covers.
result IP-generated codebooks achieve high nominal and robust adversarial accuracy.

New linear algorithms improve wSVMs for multiclass probability estimation.

problem Estimating conditional probabilities for multiclass problems.
method Proposed baseline learning and OVA learning schemes to improve wSVMs.
result Linear algorithms achieve optimal computational efficiency and good estimation accuracy.

Broker uses multi-task dynamic pricing to learn competitive prices in credit markets.

problem Lack of data and infrequent trading in credit markets.
method Two-Stage Multi-Task (TSMT) algorithm that leverages shared structure across securities.
result TSMT algorithm achieves a regret bound of O(TMd+Md)O(\sqrt{T M d} + M d), outperforming baselines.

Paper analyzes InstaHide's security, recovering all private images with provable guarantee.

problem Protecting privacy of training data in neural networks.
method Unified framework to understand and analyze attacks on InstaHide, presenting a new algorithm to recover all private images with provable guarantee.
result InstaHide is computationally secure but not information-theoretically secure when mixing two private images.

Any given classification problem can be modeled using multi-class or One-vs-All (OVA) architecture. An OVA system consists of as many OVA models as the number of classes, providing the advantage of asynchrony, where each OVA model can be re-trained independent of other models. This is particularly advantageous in setti…

2019-06-20abs ↗pdf ↗

Hybrid quantum-classical model boosts S&P 500 prediction accuracy to 60.14%.

problem Challenges in financial market prediction, especially high noise and non-stationarity.
method Combines quantum sentiment analysis, Decision Transformer, and model selection strategies.
result Achieved 60.14% directional accuracy on S&P 500, a 3.10% improvement.

Deep learning classifies autism vs controls with high accuracy using large fMRI dataset.

problem Classification difficulty of autism vs typically developing controls with fMRI data.
method Ensemble CNN model trained on 43,858 fMRI datapoints, employing class-balancing and visualization methods.
result Deep learning models achieve AUROCs of 0.6774, 0.7680, and 0.9222 for ASD vs TD, gender, and task vs rest classifications.

Paper compares Rényi min-entropy vs Shannon entropy for feature selection in machine learning.

problem Feature selection in machine learning to improve model performance.
method Proposes an algorithm based on conditional Rényi min-entropy for feature selection, comparing it to Shannon-based mutual information.
result Rényi-based algorithm tends to outperform Shannon-based in real datasets.

A common method of generalizing binary to multi-class classification is the error correcting code (ECC). ECCs may be optimized in a number of ways, for instance by making them orthogonal. Here we test two types of orthogonal ECCs on seven different datasets using three types of binary classifier and compare them with t…

2018-01-27abs ↗pdf ↗

Study reveals how initialization scale affects training accuracy in linear networks.

problem Understanding implicit bias in linear classification models.
method Asymptotic analysis of gradient flow trajectories and training loss minimization.
result Implicit bias is more complex at reasonable initialization scales and training accuracies.

New study analyzes security of neural network data reconstruction attacks.

problem Data reconstruction attacks pose a threat to private training data.
method Analyzes security boundary of data reconstruction attacks via neuron exclusivity state.
result Characterizes insecure/secure boundary of data reconstruction attacks.

Donor-aware scRNA-seq benchmarks improve classification accuracy in inflammatory bowel disease.

problem Influenza disease classification from scRNA-seq data is prone to donor-level confounding.
method Developed and evaluated three feature representations across two IBD cohorts.
result Compartment-stratified CLR composition and GatedStructuralCFN embeddings outperform linear models in classification accuracy.