Filters








4,493 Hits in 6.6 sec

On-Line Paging Against Adversarially Biased Random Inputs

Neal E. Young
2000 Journal of Algorithms  
Koutsoupias and Papadimitriou recently analyzed the least-recently-used (LRU) paging strategy in this manner, analyzing its performance on an input sequence generated by a so-called diffuse adversary -  ...  An intermediate approach is to show that an algorithm does well on a broad class of input distributions.  ...  A where A ranges over all deterministic on-line algorithms. Analogously, Ž define the optimal ratio for randomized on-line algorithms against the .  ... 
doi:10.1006/jagm.2000.1099 fatcat:k7sdohljfnerllo7ezqt6fgiye

On Guaranteed Optimal Robust Explanations for NLP Models [article]

Emanuele La Malfa, Agnieszka Zbrzezny, Rhiannon Michelmore, Nicola Paoletti, Marta Kwiatkowska
2021 arXiv   pre-print
We show how our method can be con-figured with different perturbation sets in the em-bedded space and used to detect bias in predictions by enforcing include/exclude constraints on biased terms, as well  ...  We build on abduction-based explanations for ma-chine learning and develop a method for computing local explanations for neural network models in natural language processing (NLP).  ...  Sparse Adversarial Attacks In Algorithm 2 we present a method to generate sparse adversarial attacks against features (i.e., words) of a generic input text.  ... 
arXiv:2105.03640v2 fatcat:kc7tgbnhbre6hpnncllu4gad2u

Memory-Tight Reductions [chapter]

Benedikt Auerbach, David Cash, Manuel Fersch, Eike Kiltz
2017 Lecture Notes in Computer Science  
Cryptographic reductions typically aim to be tight by transforming an adversary A into an algorithm that uses essentially the same resources as A.  ...  We argue that the amount of working memory used (relative to the initial adversary) is a relevant parameter in reductions, and that reductions that are inefficient with memory will sometimes yield less  ...  Finally, we are grateful to one of the CRYPTO 2017 reviewers for his/her very detailed and thoughtful review.  ... 
doi:10.1007/978-3-319-63688-7_4 fatcat:mhn6mmbgtbgtvb3xxn223s3iva

Page 5810 of Mathematical Reviews Vol. , Issue 2001H [page]

2001 Mathematical Reviews  
adversarially biased random inputs.  ...  The other is an NC approximation algorithm achieving an approximation ratio of 1/(2+.) for any fixed constant e > 0.” 2001h:68170 68W40 68N25 Young, Neal E. (1-DTM; Hanover, NH) On-line paging against  ... 

Dispelling Myths on Superposition Attacks: Formal Security Model and Attack Analyses [article]

Luka Music, Céline Chevalier, Elham Kashefi
2020 arXiv   pre-print
actions coherently on quantum states.  ...  It is of folkloric belief that the security of classical cryptographic protocols is automatically broken if the Adversary is allowed to perform superposition queries and the honest players forced to perform  ...  Lemma 3 (Security of One-Time-Pad against Adversaries with Superposition Access).  ... 
arXiv:2007.00677v1 fatcat:d725zyxxm5emzpsi2arzxt45q4

Towards Robust Explanations for Deep Neural Networks

Ann-Kathrin Dombrowski, Christopher J. Anders, Klaus-Robert Müller, Pan Kessel
2021 Pattern Recognition  
Based on these theoretical insights, we present three different techniques to boost robustness against manipulation: training with weight decay, smoothing activation functions, and minimizing the Hessian  ...  We develop a unified theoretical framework for deriving bounds on the maximal manipulability of a model.  ...  A simple defense against adversarial at- tacks on heatmap explanations. 2020 Workshop on Human Interpretability 545 in Machine Learning (WHI), 2020.  ... 
doi:10.1016/j.patcog.2021.108194 fatcat:qv77e5cilzfudkd2dcce3fkgsy

Reusable cryptographic fuzzy extractors

Xavier Boyen
2004 Proceedings of the 11th ACM conference on Computer and communications security - CCS '04  
As an illustration, we demonstrate how to use a biometric secret in a remote error tolerant authentication protocol that does not require any storage on the client's side.  ...  Randomness Extractors Intuitively, a (non-fuzzy) strong randomness extractor [NZ96] is a randomized function that tranforms its input from any biased distribution of sufficient min-entropy into an output  ...  This requires that there be a random execution of Fsk that on input δ 0 [w] = w outputs p 0 .  ... 
doi:10.1145/1030083.1030096 dblp:conf/ccs/Boyen04 fatcat:24mmg6dvenbcbmt4kskomwgzpm

Backdoors in Pseudorandom Number Generators: Possibility and Impossibility Results [chapter]

Jean Paul Degabriele, Kenneth G. Paterson, Jacob C. N. Schuldt, Joanne Woodage
2016 Lecture Notes in Computer Science  
In this paper, we continue the foundational line of work initiated by Dodis et al., providing both positive and negative results.  ...  Finally, and ending on a positive note, we give an impossibility result: we provide a bound on the number of previous phases that Big Brother can compromise as a function of the state-size of the generator  ...  These generate pseudorandom bits instead of truly random bits; PRNGs with input can also have their state regularly refreshed with fresh entropy, though from a possibly biased source of randomness.  ... 
doi:10.1007/978-3-662-53018-4_15 fatcat:nmu577ia2raslepa6djldqk52u

Certified Robustness via Locally Biased Randomized Smoothing

Brendon G. Anderson, Somayeh Sojoudi
2022 Conference on Learning for Dynamics & Control  
We generalize the smoothing framework to remove this assumption and learn a locally optimal robustification of the decision boundary based on training data, a method we term locally biased randomized smoothing  ...  Randomized smoothing remains one of the state-of-the-art methods for robustification with theoretical guarantees.  ...  Intuitively, this ensemble approach averages out any outlier inputs that may have drastically changed the prediction-such inputs are termed adversarial inputs or adversarial attacks.  ... 
dblp:conf/l4dc/AndersonS22 fatcat:ehcyzhpkqnff7avxakjnum65zm

Online Template Attacks: Revisited

Alejandro Cabrera Aldaya, Billy Bob Brumley
2021 Transactions on Cryptographic Hardware and Embedded Systems  
This highlights the importance of developing secure-by-default implementations, instead of fix-on-demand ones.  ...  This demonstrates that randomizing the initial targeted algorithm state does not prevent the attack as believed in previous works.We analyze three libraries libgcrypt, mbedTLS, and wolfSSL using two microarchitecture  ...  End-to-end attack on mbedTLS mbedTLS scalar multiplication (Algorithm 3) has an OTA countermeasure in place: It randomizes the starting coordinates of R just after line 3.  ... 
doi:10.46586/tches.v2021.i3.28-59 fatcat:p3v3uan37beprersu6oyuvggpi

Query-Biased Preview over Outsourced and Encrypted Data

Ningduo Peng, Guangchun Luo, Ke Qin, Aiguo Chen
2013 The Scientific World Journal  
Based on private information retrieval protocol and the core concept of searchable encryption, we propose a single-server and two-round solution to securely obtain a query-biased snippet over the encrypted  ...  An informative query-biased preview feature, as applied in modern search engine, could help the users to learn about the content without downloading the entire document.  ...  In modern search engine, if a user searches for a web page by keywords, the search engine will return the name, URI, and a small query-biased snippet for each matched page.  ... 
doi:10.1155/2013/860621 pmid:24078798 pmcid:PMC3775409 fatcat:3eyx3wwz2vewznfazbdi6os7sy

Adversarial Network Embedding [article]

Quanyu Dai, Qiang Li, Jian Tang, Dan Wang
2017 arXiv   pre-print
As shown by the empirical results, our method is competitive with or superior to state-of-the-art approaches on benchmark network embedding tasks.  ...  Specifically, we propose an Adversarial Network Embedding (ANE) framework, which leverages the adversarial learning principle to regularize the representation learning.  ...  One simple way is to directly use existing scalable methods, e.g. DeepWalk and LINE, to obtain initial embeddings X as input. https://www.csie.ntu.edu.tw/ cjlin/liblinear/  ... 
arXiv:1711.07838v1 fatcat:urnmryjidfgr3di4znm6dh2p2u

Towards Context-Agnostic Learning Using Synthetic Data [article]

Charles Jin, Martin Rinard
2021 arXiv   pre-print
We propose a novel setting for learning, where the input domain is the image of a map defined on the product of two sets, one of which completely determines the labels.  ...  on real world data without our techniques yields classifiers that are brittle to perturbations of the background.  ...  Adversarial discriminative domain adaptation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 7167–7176, 2017. Stéfan van der Walt, Johannes L.  ... 
arXiv:2005.14707v3 fatcat:3lnb7rzqcjeh7jldywsvw3u2zi

Has CEO Gender Bias Really Been Fixed? Adversarial Attacking and Improving Gender Fairness in Image Search

Yunhe Feng, Chirag Shah
2022 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Gender bias is one of the most common and well-studied demographic biases in information retrieval, and in general in AI systems.  ...  In this paper, we propose adversarial attack queries composed of professions and countries (e.g., 'CEO United States') to investigate whether gender bias is thoroughly mitigated by image search engines  ...  We must also look at inherent disparity and biases carried out in them. Among different types of image biases, gender bias is one of the most common and well-studied demographic biases.  ... 
doi:10.1609/aaai.v36i11.21445 fatcat:t34b72rntvg3lodyh64cvcdiji

Malicious PDF Detection Model against Adversarial Attack Built from Benign PDF Containing JavaScript

Ah Reum Kang, Young-Seob Jeong, Se Lyeong Kim, Jiyoung Woo
2019 Applied Sciences  
We found that random forest, an ensemble algorithm of a decision tree, exhibits a good performance on malware detection and is robust for adversarial samples.  ...  We build the adversarial samples by injecting the malware codes into base samples. The proposed model is evaluated against a large collection of malicious and benign PDFs.  ...  The random forest algorithm has strong generalization ability and strong robustness against data noise.  ... 
doi:10.3390/app9224764 fatcat:zr7tded7ibduvllyl3wicazwf4
« Previous Showing results 1 — 15 out of 4,493 results