4,031 Hits in 6.0 sec

Training Language Models under Resource Constraints for Adversarial Advertisement Detection

Eshwar Shamanna Girishekar, Shiv Surya, Nishant Nikhil, Dyut Kumar Sil, Sumit Negi, Aruna Rajan
2021 Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Industry Papers   unpublished
This paper focusses on techniques for training text classification models under resource constraints, built as part of automated solutions for advertising content moderation.  ...  Our extensive experiments on multiple languages show that these techniques detect adversarial ad categories with a substantial gain in precision at high recall threshold over the baseline.  ...  Our work focuses on techniques we leverage to train state of the art language models for detecting adversarial advertising content in text.  ... 
doi:10.18653/v1/2021.naacl-industry.35 fatcat:b2x7l3yme5dctkzdwsxb7bhpcy

Tracking Cyber Adversaries with Adaptive Indicators of Compromise [article]

Justin E. Doak, Joe B. Ingram, Sam A. Mulder, John H. Naegle, Jonathan A. Cox, James B. Aimone, Kevin R. Dixon, Conrad D. James, David R. Follett
2017 arXiv   pre-print
This result suggests that self-updating the model over time should be done in practice to continue to detect known, evolving adversaries.  ...  However, the difference in overall detection performance, as measured by the area under the curve (AUC), between the two methods is negligible.  ...  ., for the U.S. Department of Energy's National Nuclear Security Administration under contract de-na0003525.  ... 
arXiv:1712.07671v1 fatcat:wva6jbjei5cchdokfydzssflje

Code-Mixing on Sesame Street: Dawn of the Adversarial Polyglots [article]

Samson Tan, Shafiq Joty
2021 arXiv   pre-print
Finally, we propose an efficient adversarial training scheme that trains in the same number of steps as the original model and show that it improves model accuracy.  ...  Inspired by this phenomenon, we present two strong black-box adversarial attacks (one word-level, one phrase-level) for multilingual models that push their ability to handle code-mixed sentences to the  ...  BERT: Pre-training of deep bidirectional transformers for language under- standing.  ... 
arXiv:2103.09593v3 fatcat:epgdk4dr3zg7bn5jjqpaediwzy

Adversarial Attacks and Defense Technologies on Autonomous Vehicles: A Review

K. T. Y. Mahima, Mohamed Ayoob, Guhanathan Poravi
2021 Applied Computer Systems  
For most of these components, computer vision technologies with deep learning such as object detection, semantic segmentation and image classification are used.  ...  Therefore, implementing defense models against adversarial attacks has become an increasingly critical research area.  ...  A 2017 [58] To evaluate the robustness of the adversarial perturbations to traffic sign object detectors under the physical world constraints.  ... 
doi:10.2478/acss-2021-0012 fatcat:runxr47gzrb6znc4hygld36qgy

Machine Learning Methods for Computer Security (Dagstuhl Perspectives Workshop 12371)

Anthony D. Joseph, Pavel Laskov, Fabio Roli, J. Doug Tygar, Blaine Nelson, Marc Herbstritt
2013 Dagstuhl Reports  
The second group focused on the current approaches and methodical challenges for learning in security-sensitive adversarial domains.  ...  Learning-based approaches are particularly advantageous for security applications designed to counter sophisticated and evolving adversaries because learning methods can cope with large amounts of evolving  ...  What are the theoretical limitations of worst-case attacks against learning algorithms under different constraints?  ... 
doi:10.4230/dagrep.2.9.109 dblp:journals/dagstuhl-reports/JosephLRTN12 fatcat:4x3ng2szxfg5jnkf5rtwsmttrm

A Review on Android Malware: Attacks, Countermeasures and Challenges Ahead

ShymalaGowri Selvaganapathy, Sudha Sadasivam, Vinayakumar Ravi
2021 Journal of Cyber Security and Mobility  
detection phase.  ...  It sheds spotlight for researchers to perceive the current state of the art techniques available to fend off malware along with suggestions on possible future directions  ...  They perform offline detection of malware with no constraints in regard to computing capabilities or handling huge scale of samples for data processing when compared to the energy and computation resources  ... 
doi:10.13052/jcsm2245-1439.1017 fatcat:mtxfys7pwvb7dastdlyu2s2tzq

Threats to Federated Learning: A Survey [article]

Lingjuan Lyu, Han Yu, Qiang Yang
2020 arXiv   pre-print
With the emergence of data silos and popular privacy awareness, the traditional centralized approach of training artificial intelligence (AI) models is facing strong challenges.  ...  Federated learning (FL) has recently emerged as a promising solution under this new reality.  ...  In addition, adversaries must be selected for many rounds of FL training. Thus, it is not suitable for H2C scenarios, but more likely under H2B scenarios.  ... 
arXiv:2003.02133v1 fatcat:htv4tztwlbdihdkat5bzlcm46y

Federated Learning Challenges and Opportunities: An Outlook [article]

Jie Ding, Eric Tramel, Anit Kumar Sahu, Shuang Wu, Salman Avestimehr, Tao Zhang
2022 arXiv   pre-print
Although many methods and applications have been developed for FL, several critical challenges for practical FL systems remain unaddressed.  ...  Federated learning (FL) has been developed as a promising framework to leverage the resources of edge devices, enhance customers' privacy, comply with regulations, and reduce development costs.  ...  First, under memory constraints, each on-device model needs to be small in size. This poses a methodological challenge for the server to leverage a large function class for large data (in hindsight).  ... 
arXiv:2202.00807v1 fatcat:aftvlmxw3zcjvpkcaaj22boe2e

A Cross-Media Advertising Design and Communication Model Based on Feature Subspace Learning

Shanshan Li, Gengxin Sun
2022 Computational Intelligence and Neuroscience  
This paper uses feature subspace learning and cross-media retrieval analysis to construct an advertising design and communication model.  ...  Through the study of cross-media advertising design and communication models based on feature subspace learning, it is of positive significance to advance commercial advertising design by guiding designers  ...  Cross-Media Model Design for Feature Subspace Learning From the existing feature subspace learning models, most forms of constraint terms have contributed to the learning of feature subspaces.  ... 
doi:10.1155/2022/5874722 pmid:35619757 pmcid:PMC9129948 fatcat:qsh45riw7vfhfjrvorwwamf7ne

A Survey on Resilient Machine Learning [article]

Atul Kumar, Sameep Mehta
2017 arXiv   pre-print
However, recent research has shown that machine learning models are venerable to attacks by adversaries at all phases of machine learning (eg, training data collection, training, operation).  ...  Machine learning based system are increasingly being used for sensitive tasks such as security surveillance, guiding autonomous vehicle, taking investment decisions, detecting and blocking network intrusion  ...  Network Intrusion Detection For intrusion detection, statistical machine learning based techniques build a model for normal behavior from training data and then detect attacks that deviates from that model  ... 
arXiv:1707.03184v1 fatcat:qjylw7bvkzbdlbrof5cfpy2jyq

A Survey on Bias and Fairness in Machine Learning [article]

Ninareh Mehrabi, Fred Morstatter, Nripsuta Saxena, Kristina Lerman, Aram Galstyan
2022 arXiv   pre-print
We have recently seen work in machine learning, natural language processing, and deep learning that addresses such challenges in different subdomains.  ...  We then created a taxonomy for fairness definitions that machine learning researchers have defined in order to avoid the existing bias in AI systems.  ...  Language Model.  ... 
arXiv:1908.09635v3 fatcat:fygrqs3sing6zdsg53t7awhih4

Review of Android Malware Detection Based on Deep Learning

Zhiqiang Wang, Qian Liu, Yaping Chi
2020 IEEE Access  
architecture and detection schemes, and analyzing existing problems and challenges.  ...  It has posed a severe threat to cyberspace security because traditional detection methods have many limitations.  ...  We gratefully acknowledge the anonymous reviewers for their valuable comments.  ... 
doi:10.1109/access.2020.3028370 fatcat:tujn3ghssrfkffzafat7l3cnse

Privacy Protection of Grid Users Data with Blockchain and Adversarial Machine Learning [article]

Ibrahim Yilmaz, Kavish Kapoor, Ambareen Siraj, Mahmoud Abouyoussef
2021 arXiv   pre-print
We then introduce Adversarial Machine Learning Occupancy Detection Avoidance with Blockchain (AMLODA-B) framework as a counter-attack by deploying an algorithm based on the Long Short Term Memory (LSTM  ...  ) model into the standardized smart metering infrastructure to prevent leakage of consumers personal information.  ...  ACKNOWLEDGMENT This work is funded by CESR (Center for Energy Systems Research) at Tennessee Technological University with resource support from the Cybersecurity Education Research and Outreach Center  ... 
arXiv:2101.06308v1 fatcat:j4fwirwtqvbfvmr7em7kyac2oa

A review of attacks and security approaches in open multi-agent systems

Shahriar Bijani, David Robertson
2012 Artificial Intelligence Review  
Focusing on information leakage in choreography systems using LCC, we then suggest two frameworks to detect insecure information flows: conceptual modeling of interaction models and languagebased information  ...  A major practical limitation to such systems is security, because the very openness of such systems opens the doors to adversaries for exploit existing vulnerabilities.  ...  Anomalies are detected when the current MAS state differs from the trained model (classifier). The openness in open MASs might hinder successful anomaly detection.  ... 
doi:10.1007/s10462-012-9343-1 fatcat:ppins5fil5hztjftmzbgzknzli

Adversarial machine learning

Ling Huang, Anthony D. Joseph, Blaine Nelson, Benjamin I.P. Rubinstein, J. D. Tygar
2011 Proceedings of the 4th ACM workshop on Security and artificial intelligence - AISec '11  
for modeling an adversary's capabilities; explore the limits of an adversary's knowledge about the algorithm, feature space, training, and input data; explore vulnerabilities in machine learning algorithms  ...  In this paper, we: give a taxonomy for classifying attacks against online machine learning algorithms; discuss application-specific factors that limit an adversary's capabilities; introduce two models  ...  many fruitful discussions and collaborations that have influenced our thinking about adversarial machine learning.  ... 
doi:10.1145/2046684.2046692 dblp:conf/ccs/HuangJNRT11 fatcat:d6wcto4tmvbbrec35cjdengxby
« Previous Showing results 1 — 15 out of 4,031 results