Filters








215 Hits in 4.5 sec

Purifying data by machine learning with certainty levels

Shlomi Dolev, Guy Leshem, Reuven Yagel
2010 Proceedings of the Third International Workshop on Reliability, Availability, and Security - WRAS '10  
This paper introduces the use of a certainty level measure to obtain better classification capability in the presence of corrupted data items.  ...  This method is of independent interest, that of significantly improving the classification of the random forest machine learning technique in less severe settings.  ...  In the Probably Approximately Correct (PAC) learning framework, Valiant (Valiant, 1984) introduced the notion of PAC learning in the presence of malicious noise.  ... 
doi:10.1145/1953563.1953567 dblp:conf/podc/DolevLY10 fatcat:5tsho3pw6vcdddhkvxp2ykupjm

Page 5392 of Mathematical Reviews Vol. , Issue 94i [page]

1994 Mathematical Reviews  
BELL; Murray Hill, NJ); Li, Ming [Li, Ming"] (3-WTRL-C; Waterloo, ON) Learning in the presence of malicious errors.  ...  ThePAC” model of learning was proposed by L. G. Valiant [Comm.  ... 

Learning Logic programs with random classification noise [chapter]

Tamás Horváth, Robert H. Sloan, György Turán
1997 Lecture Notes in Computer Science  
We review the polynomial PAC learnability of nonrecursive, determinate, constant-depth Horn clauses in the presence of such noise.  ...  Also, we show that arbitrary nonrecursive Horn clauses with forest background knowledge remain polynomially PAC learnable in the presence of noise.  ...  Learning Conjunctions from Noisy Data Angluin and Laird 1] gave the rst algorithm for PAC learning conjunctions from data with random classi cation noise.  ... 
doi:10.1007/3-540-63494-0_63 fatcat:ecttkvrzo5eytjdjnpidqai5lm

Federated Learning from Small Datasets [article]

Michael Kamp
2021 arXiv   pre-print
This enables applications of machine learning in settings of inherently distributed, undisclosable data such as in the medical domain.  ...  Often, however, local datasets are so small that local objectives differ greatly from the global objective, resulting in federated learning to fail.  ...  Similarly, the malicious client can infer upon the presence of data points with certain attributes in the dataset (Ateniese et al., 2015) .  ... 
arXiv:2110.03469v2 fatcat:7ucgfn2qqbeb3dartwl3aic3bi

On the Sample Complexity of Adversarial Multi-Source PAC Learning [article]

Nikola Konstantinov, Elias Frantar, Dan Alistarh, Christoph H. Lampert
2020 arXiv   pre-print
We study the problem of learning from multiple untrusted data sources, a scenario of increasing practical relevance given the recent emergence of crowdsourcing and collaborative learning paradigms.  ...  It is known that in the single-source case, an adversary with the power to corrupt a fixed fraction of the training data can prevent PAC-learnability, that is, even in the limit of infinitely much training  ...  Acknowledgements Dan Alistarh is supported in part by the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No 805223 ScaleML).  ... 
arXiv:2002.10384v2 fatcat:sd4daitwevhnxmcgyra3kravhu

On learning visual concepts and DNF formulae

Eyal Kushilevitz, Dan Roth
1996 Machine Learning  
We consider the problem of learning DNF formulae in the mistake-bound and the PAC models.  ...  Unlike previous learnability results for DNF (and CNF) formulae, these subclasses are not limited in the number of terms or in the number of variables per term; yet, they contain the subclasses of k-DNF  ...  We thank the referees for helpful suggestions that improved the presentation of this work. Eyal Kushilevitz was supported by research contracts ONR-N0001491-J-1981 and NSF-CCR-90-07677.  ... 
doi:10.1007/bf00117833 fatcat:vdnxtlfmwbbynk77npn4lmocm4

On learning visual concepts and DNF formulae

Eyal Kushilevitz, Dan Roth
1993 Proceedings of the sixth annual conference on Computational learning theory - COLT '93  
We consider the problem of learning DNF formulae in the mistake-bound and the PAC models.  ...  Unlike previous learnability results for DNF (and CNF) formulae, these subclasses are not limited in the number of terms or in the number of variables per term; yet, they contain the subclasses of k-DNF  ...  We thank the referees for helpful suggestions that improved the presentation of this work. Eyal Kushilevitz was supported by research contracts ONR-N0001491-J-1981 and NSF-CCR-90-07677.  ... 
doi:10.1145/168304.168362 dblp:conf/colt/KushilevitzR93 fatcat:4ivza5tdqrgyvcm7obfq4clayi

Open problems in the security of learning

Marco Barreno, Peter L. Bartlett, Fuching Jack Chi, Anthony D. Joseph, Blaine Nelson, Benjamin I.P. Rubinstein, Udam Saini, J. D. Tygar
2008 Proceedings of the 1st ACM workshop on Workshop on AISec - AISec '08  
In this paper, we present three broad research directions towards the end of developing truly secure learning.  ...  Machine learning has become a valuable tool for detecting and preventing malicious activity.  ...  This work was supported in part by the Team for Research in Ubiquitous Secure Technology (TRUST), which receives support from the National Science Foundation (  ... 
doi:10.1145/1456377.1456382 dblp:conf/ccs/BarrenoBCJNRST08 fatcat:4uk7kufh4zevfgxkvhz7t4qvm4

Table of contents

2019 2019 IEEE 8th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP)  
Moura, Carnegie Mellon University, United States MA1.5: PAC LEARNING FROM DISTRIBUTED DATA IN THE PRESENCE OF ...............................186 MALICIOUS NODES Zhixiong Yang, Waheed Bajwa, Rutgers University-New  ...  Brunswick, United States TM1: Learning over Graphs TM1.1: LEARNING SPARSE HYPERGRAPHS FROM DYADIC RELATIONAL DATA ......................216 Mario Coutino, TU Delft, Netherlands; Sundeep Chepuri, Indian  ...  Joaquin Míguez, Universidad Carlos III de Madrid, Spain; Lucas Lacasa, Queen Mary University of London, United Kingdom; Universidad Carlos III de Madrid, Spain; Inés P.  ... 
doi:10.1109/camsap45676.2019.9022485 fatcat:eiz3bmwzdrfojc36xlrzbrppdq

Fog computing security: a review of current applications and security solutions

Saad Khan, Simon Parkinson, Yongrui Qin
2017 Journal of Cloud Computing: Advances, Systems and Applications  
In addition, Fog systems are capable of processing large amounts of data locally, operate on-premise, are fully portable, and can be installed on heterogeneous hardware.  ...  For example, Internet of Things (IoT) devices are required to quickly process a large amount of data.  ...  However, the presence of such private data puts the Fog platform in a sensitive position.  ... 
doi:10.1186/s13677-017-0090-3 fatcat:g3q3q3ctf5coteaycnceib522u

Knowing what doesn't matter: exploiting the omission of irrelevant data

Russell Greiner, Adam J. Grove, Alexander Kogan
1997 Artificial Intelligence  
In many diagnostic tasks, however, the data will include the values of only some of the attributes; we model this as a blocking process that hides the values of those attributes from the learner.  ...  PAC model-viz., decision trees and DNF formulae-are trivial to learn in this setting.  ...  PAC-learning decision trees in the presence of classification noise follows from the techniques of the previous theorem. Namely, first let k > 8/( 1 -2a)* ln(2s/6) (which improves on Eq.  ... 
doi:10.1016/s0004-3702(97)00048-9 fatcat:5ldpntwiyje6rhymbd6h476li4

Cybersecurity in PACS and Medical Imaging: an Overview

Marco Eichelberg, Klaus Kleber, Marc Kämmerer
2020 Journal of digital imaging  
From a practical perspective, PACS specific security measures must be implemented together with the measures applicable to the IT infrastructure as a whole, in order to prevent incidents such as PACS systems  ...  The article concludes with a discussion of gaps in the body of published literature and a summary.  ...  Eichelberg reports a grant from VISUS Health IT, during the conduct of the study. Mr. Kleber and Dr.  ... 
doi:10.1007/s10278-020-00393-3 pmid:33123867 fatcat:6cf63xnkr5feznvca5e6a3joqe

A neuroidal architecture for cognitive computation [chapter]

Leslie G. Valiant
1998 Lecture Notes in Computer Science  
Attribute-efficient learning algorithms, which allow learning from few examples in large dimensional systems, are fundamental to the approach.  ...  Its aim is to exploit to the full those aspects of computational learning that are known to offer powerful solutions in the acquisition and maintenance of robust knowledge bases.  ...  I am grateful to Roni Khardon, Dan Roth, and Rocco Servedio and several referees for their helpful comments on various versions of this paper.  ... 
doi:10.1007/bfb0055091 fatcat:b6534lxzkjcdrffkcvy25wiy6q

A neuroidal architecture for cognitive computation

Leslie G. Valiant
2000 Journal of the ACM  
Attribute-efficient learning algorithms, which allow learning from few examples in large dimensional systems, are fundamental to the approach.  ...  Its aim is to exploit to the full those aspects of computational learning that are known to offer powerful solutions in the acquisition and maintenance of robust knowledge bases.  ...  I am grateful to Roni Khardon, Dan Roth, and Rocco Servedio and several referees for their helpful comments on various versions of this paper.  ... 
doi:10.1145/355483.355486 fatcat:ydvy6dok6ra2vcbnuqngi6fiey

Mechanism Design for Cost Optimal PAC Learning in the Presence of Strategic Noisy Annotators [article]

Dinesh Garg, Sourangshu Bhattacharya, S. Sundararajan, Shirish Shevade
2012 arXiv   pre-print
We consider the problem of Probably Approximate Correct (PAC) learning of a binary classifier from noisy labeled examples acquired from multiple annotators (each characterized by a respective classification  ...  For this scenario, we design a cost optimal procurement auction mechanism along the lines of Myerson's optimal auction design framework in a non-trivial manner.  ...  this algorithm in the presence of n noisy annotators.  ... 
arXiv:1210.4859v1 fatcat:ajlacnzry5f4lnfdudf4ndcjf4
« Previous Showing results 1 — 15 out of 215 results