28,736 Hits in 4.4 sec

Differentially Private Exploration in Reinforcement Learning with Linear Representation [article]

Paul Luyo and Evrard Garcelon and Alessandro Lazaric and Matteo Pirotta
2021 arXiv   pre-print
This paper studies privacy-preserving exploration in Markov Decision Processes (MDPs) with linear representation.  ...  We first consider the setting of linear-mixture MDPs (Ayoub et al., 2020) (a.k.a. model-based setting) and provide an unified framework for analyzing joint and local differential private (DP) exploration  ...  In this paper, we contribute to the study of DP in online reinforcement learning (RL).  ... 
arXiv:2112.01585v2 fatcat:ubnlne4zyrgqrfcb7gcyhzznt4

Fair NLP Models with Differentially Private Text Encoders [article]

Gaurav Maheshwari, Pascal Denis, Mikaela Keller, Aurélien Bellet
2022 arXiv   pre-print
In this work, we propose FEDERATE, an approach that combines ideas from differential privacy and adversarial training to learn private text representations which also induces fairer models.  ...  Our results show that FEDERATE consistently improves upon previous methods, and thus suggest that privacy and fairness can positively reinforce each other.  ...  encoder to get private representations of input texts, and is further combined with adversarial training to learn fair models.  ... 
arXiv:2205.06135v1 fatcat:j33p3kqv6fgfdm5fgcrnjeqmsa

Discovering Differential Features: Adversarial Learning for Information Credibility Evaluation [article]

Lianwei Wu, Yuan Rao, Ambreen Nazir, Haolin Jin
2019 arXiv   pre-print
and KL-divergence for making the private features more differential.  ...  Specifically, ANSP involves two tasks: one is to prevent the binary classification of true and false information for capturing common features relying on adversarial networks guided by reinforcement learning  ...  Reinforcement Learning (RL) With the fame of AlphaGo, reinforcement learning has become more and more popular in academic communities.  ... 
arXiv:1909.07523v1 fatcat:brv5rqpqcfbhrgny3hqrzws5qm

Branch Prediction as a Reinforcement Learning Problem: Why, How and Case Studies [article]

Anastasios Zouzias, Kleovoulos Kalaitzidis, Boris Grot
2021 arXiv   pre-print
This paper argues that looking at BP from the viewpoint of Reinforcement Learning (RL) facilitates systematic reasoning about, and exploration of, BP designs.  ...  Recent years have seen stagnating improvements to branch predictor (BP) efficacy and a dearth of fresh ideas in branch predictor design, calling for fresh thinking in this area.  ...  REINFORCEMENT LEARNING BACKGROUND Reinforcement learning algorithms are well-suited for scenarios where an agent can learn from an environment.  ... 
arXiv:2106.13429v1 fatcat:qjhqs2eqhjhmdodfas5nh4e3r4

Federated Learning System without Model Sharing through Integration of Dimensional Reduced Data Representations [article]

Anna Bogdanova, Akie Nakai, Yukihiko Okada, Akira Imakura, Tetsuya Sakurai
2020 arXiv   pre-print
In this work, we explore an alternative federated learning system that enables integration of dimensionality reduced representations of distributed data prior to a supervised learning task, thus avoiding  ...  Dimensionality Reduction is a commonly used element in a machine learning pipeline that helps to extract important features from high-dimensional data.  ...  Acknowledgments The present study is supported in part by the Japan Science and Technology Agency (JST), ACT-I (No.  ... 
arXiv:2011.06803v1 fatcat:7klksw55fzadjjldu4v7aktfv4

Deep Reinforcement Learning-based Text Anonymization against Private-Attribute Inference

Ahmadreza Mosallanezhad, Ghazaleh Beigi, Huan Liu
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
In this paper, we study the problem of textual data anonymization and propose a novel Reinforcement Learning-based Text Anonymizor, RLTA, which addresses the problem of private-attribute leakage while  ...  Our approach first extracts a latent representation of the original text w.r.t. a given task, then leverages deep reinforcement learning to automatically learn an optimal strategy for manipulating text  ...  This material is based upon the work supported, in part, by NSF 1614576, ARO W911NF-15-1-0328 and ONR N00014-17-1-2605.  ... 
doi:10.18653/v1/d19-1240 dblp:conf/emnlp/MosallanezhadBL19 fatcat:wld7k4kt2vhe3h6pmf76u7ftny

Deep Reinforcement Learning [article]

Yuxi Li
2018 arXiv   pre-print
We discuss deep reinforcement learning in an overview style. We draw a big picture, filled with details.  ...  We start with background of artificial intelligence, machine learning, deep learning, and reinforcement learning (RL), with resources.  ...  Lanctot et al. (2017) observe that independent RL, in which each agent learns by interacting with the environment, oblivious to other agents, can overfit the learned policies to other agents' policies  ... 
arXiv:1810.06339v1 fatcat:kp7atz5pdbeqta352e6b3nmuhy

Variational Bayes in Private Settings (VIPS) (Extended Abstract)

James R. Foulds, Mijung Park, Kamalika Chaudhuri, Max Welling
2020 Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence  
learning.  ...  Our framework respects differential privacy, the gold-standard privacy criterion.  ...  On Reinforcement Learning for GM Using reinforcement learning (RL) for GM is still in its infancy.  ... 
doi:10.24963/ijcai.2020/694 dblp:conf/ijcai/YanYH20 fatcat:pc4nelo7gzfmvmsiym3ohwspxa

Spatial navigation model based on chaotic attractor networks

Horatiu Voicu, Robert Kozma, Derek Wong, Walter J. Freeman
2004 Connection science  
The system includes a hippocampal module that processes global spatial information and a cortical module that deals with local sensory information.  ...  We use Hebbian learning in combination with reinforcement. Learning occurs if the system receives positive or negative reinforcement.  ...  The impulse response was used to identify the parameters of a second-order linear differential equation that describes the behaviour of neuronal populations.  ... 
doi:10.1080/09540090410001664641 fatcat:bkijao4ly5clzpmees2hshc7j4

Differentially Private Graph Classification with GNNs [article]

Tamara T. Mueller, Johannes C. Paetzold, Chinmay Prabhakar, Dmitrii Usynin, Daniel Rueckert, Georgios Kaissis
2022 arXiv   pre-print
Finally, we apply explainability techniques to assess whether similar representations are learned in the private and non-private settings and establish robust baselines for future work in this area.  ...  Machine learning with differential privacy is a promising technique to allow deriving insight from sensitive data while offering formal guarantees of privacy protection.  ...  Different approaches to the here introduced application of differential privacy have been explored in the context of federated learning on graphs and locally private graph neural network training.  ... 
arXiv:2202.02575v2 fatcat:xuozco5iy5apvhp4mzhkbvcdzi

Exploring Shared Structures and Hierarchies for Multiple NLP Tasks [article]

Junkun Chen and Kaiyu Chen and Xinchi Chen and Xipeng Qiu and Xuanjing Huang
2018 arXiv   pre-print
The controller is trained with reinforcement learning to maximize the expected accuracies for all tasks.  ...  Designing shared neural architecture plays an important role in multi-task learning.  ...  Thus, we can use reinforcement learning to train the controller.  ... 
arXiv:1808.07658v1 fatcat:vwj3ucvf4rdu5ocvflmc454aj4

Emerging Trends in Federated Learning: From Model Fusion to Federated X Learning [article]

Shaoxiong Ji and Teemu Saravirta and Shirui Pan and Guodong Long and Anwar Walid
2021 arXiv   pre-print
We conduct a focused survey of federated learning in conjunction with other learning algorithms.  ...  transfer learning, unsupervised learning, and reinforcement learning.  ...  [39] proposed differentially private federated generative models to address the challenges of non-inspectable data scenario.  ... 
arXiv:2102.12920v2 fatcat:5fcwfhxibbedbcbuzrfyqdedky

A Survey of Adversarial Machine Learning in Cyber Warfare

Vasisht Duddu
2018 Defence Science Journal  
We explore the threat models for Machine Learning systems and describe the various techniques to attack and defend them.  ...  We present privacy issues in these models and describe a cyber-warfare test-bed to test the effectiveness of the various attack-defence strategies and conclude with some open problems in this area of research  ...  Differential privacy has been explored to ensure privacy guarantees for ML models for non-convex objective functions using differentially private stochastic gradient descent 59,61 .  ... 
doi:10.14429/dsj.68.12371 fatcat:vyupcxe6hrhllb4rowequxrf5i

Privacy-Preserving Reinforcement Learning Beyond Expectation [article]

Arezoo Rajabi, Bhaskar Ramasubramanian, Abdullah Al Maruf, Radha Poovendran
2022 arXiv   pre-print
We incorporate cumulative prospect theory (CPT) into the objective of a reinforcement learning (RL) problem for the former. For the latter, we use differential privacy.  ...  Through empirical evaluations, we highlight a privacy-utility tradeoff and demonstrate that the RL agent is able to learn behaviors that are aligned with that of a human user in the same environment in  ...  RELATED WORK This section summarizes related work in risk-sensitive and differentially private reinforcement learning.  ... 
arXiv:2203.10165v1 fatcat:7lfsu5spzfdcbkremfypvg6nw4

Cooperative Multi-agent Control Using Deep Reinforcement Learning [chapter]

Jayesh K. Gupta, Maxim Egorov, Mykel Kochenderfer
2017 Lecture Notes in Computer Science  
Using deep reinforcement learning with a curriculum learning scheme, our approach can solve problems that were previously considered intractable by most multi-agent reinforcement learning algorithms.  ...  This work considers the problem of learning cooperative policies in complex, partially observable domains without explicit communication.  ...  However, we found the reactive policy learned with an MLP representation to be unsatisfactory.  ... 
doi:10.1007/978-3-319-71682-4_5 fatcat:ie4vvneipjgxbdwngj3bncs6eu
« Previous Showing results 1 — 15 out of 28,736 results