no code implementations • 25 Apr 2024 • Yunfei Ge, Quanyan Zhu
The pervasive integration of Artificial Intelligence (AI) has introduced complex challenges in the responsibility and accountability in the event of incidents involving AI-enabled systems.
no code implementations • 1 Apr 2024 • Quanyan Zhu
This chapter starts with a systemic view toward cyber risks and presents the confluence of game theory, control theory, and learning theories, which are three major pillars for the design of cyber resilience mechanisms to counteract increasingly sophisticated and evolving threats in our networks and organizations.
no code implementations • 14 Mar 2024 • Tao Li, Quanyan Zhu
This chapter concludes with a discussion of the challenges associated with FMs and their application in the domain of cybersecurity.
no code implementations • 10 Mar 2024 • Quanyan Zhu, Tamer Basar
The article concludes by discussing the interplay between robustness and resilience, suggesting that a comprehensive theory of resilience and quantification metrics, and formalization through game-theoretic frameworks are necessary.
no code implementations • 29 Feb 2024 • Tao Li, Kim Hammar, Rolf Stadler, Quanyan Zhu
To address these limitations, we propose conjectural online learning (\textsc{col}), an online method for generic \textsc{aisg}s. \textsc{col} uses a forecaster-actor-critic (\textsc{fac}) architecture where subjective forecasts are used to conjecture the opponents' strategies within a lookahead horizon, and Bayesian learning is used to calibrate the conjectures.
1 code implementation • 19 Feb 2024 • Kim Hammar, Tao Li, Rolf Stadler, Quanyan Zhu
We study automated security response for an IT infrastructure and formulate the interaction between an attacker and a defender as a partially observed, non-stationary game.
no code implementations • 3 Jan 2024 • Yuhan Zhao, Juntao Chen, Quanyan Zhu
The attacker aims to exploit this vulnerability to enable a successful physical compromise, while the system operator's goal is to ensure a normal operation of the grid by mitigating cyber risks.
no code implementations • 6 Oct 2023 • Tao Li, Juan Guevara, Xinghong Xie, Quanyan Zhu
In the multi-agent RL (MARL) setting, this distribution shift may arise from the nonstationary opponents (exogenous agents beyond control) in the online testing who display distinct behaviors from those recorded in the offline dataset.
no code implementations • 5 Sep 2023 • Haozhe Lei, Quanyan Zhu
In the area of learning-driven artificial intelligence advancement, the integration of machine learning (ML) into self-driving (SD) technology stands as an impressive engineering feat.
no code implementations • 23 Jun 2023 • Yunian Pan, Tao Li, Henger Li, Tianyi Xu, Zizhan Zheng, Quanyan Zhu
Previous research has shown that federated learning (FL) systems are exposed to an array of security risks.
1 code implementation • 11 Jun 2023 • Mingsheng Yin, Tao Li, Haozhe Lei, Yaqi Hu, Sundeep Rangan, Quanyan Zhu
To equip the navigation agent with sample-efficient learning and {zero-shot} generalization, this work proposes a novel physics-informed RL (PIRL) where a distance-to-target-based cost (standard in e2e) is augmented with physics-informed reward shaping.
no code implementations • 1 Jun 2023 • Yunfei Ge, Quanyan Zhu
As an economic solution to compensate for potential damages, AI liability insurance is a promising market to enhance the integration of AI into daily life.
no code implementations • 3 Apr 2023 • Yunian Pan, Tao Li, Quanyan Zhu
\textit{Intelligent Navigation Systems} (INS) are exposed to an increasing number of informational attack vectors, which often intercept through the communication channels between the INS and the transportation network during the data collecting process.
no code implementations • 21 Mar 2023 • Yinan Hu, Quanyan Zhu
Human-sensor systems have a wide range of applications in fields such as robotics, healthcare, and finance.
no code implementations • 6 Mar 2023 • Yunfei Ge, Tao Li, Quanyan Zhu
The increasing connectivity and intricate remote access environment have made traditional perimeter-based network defense vulnerable.
no code implementations • 26 Jan 2023 • Yurid Nugraha, Ahmet Cetinkaya, Tomohisa Hayakawa, Hideaki Ishii, Quanyan Zhu
Consensus and cluster forming of multiagent systems in the face of jamming attacks along with reactive recovery actions by a defender are discussed.
no code implementations • 11 Jan 2023 • Juntao Chen, Junaid Farooq, Quanyan Zhu
The contract design creates a pricing structure for on-demand sensing data for IoT users.
no code implementations • 9 Jan 2023 • Yurid Nugraha, Ahmet Cetinkaya, Tomohisa Hayakawa, Hideaki Ishii, Quanyan Zhu
A two-player game-theoretic problem on resilient graphs in a multiagent consensus setting is formulated.
no code implementations • 22 Dec 2022 • Tao Li, Quanyan Zhu
This work considers a double-sided information asymmetry in a Bayesian Stackelberg game, where the leader's realized action, sampled from the mixed strategy commitment, is hidden from the follower.
no code implementations • 17 Dec 2022 • Haozhe Lei, Quanyan Zhu
To ensure traffic safety in self-driving environments and respond to vehicle-human interaction challenges such as jaywalking, we propose Level-$k$ Meta Reinforcement Learning (LK-MRL) algorithm.
no code implementations • 3 Dec 2022 • Yurid Nugraha, Tomohisa Hayakawa, Hideaki Ishii, Ahmet Cetinkaya, Quanyan Zhu
Evolution of agents' dynamics of multiagent systems under consensus protocol in the face of jamming attacks is discussed, where centralized parties are able to influence the control signals of the agents.
no code implementations • 11 Nov 2022 • Yuhan Zhao, Quanyan Zhu
To this end, we develop a meta-learning-based Stackelberg game-theoretic framework to address the challenges in the guided cooperative control for linear systems.
no code implementations • 4 Nov 2022 • Yinan Hu, Quanyan Zhu
The detection and discrimination of quantum states serve a crucial role in quantum signal processing, a discipline that studies methods and techniques to process signals that obey the quantum mechanics frameworks.
no code implementations • 6 Oct 2022 • Yunian Pan, Tao Li, Quanyan Zhu
We investigate the resilience of learning-based \textit{Intelligent Navigation Systems} (INS) to informational flow attacks, which exploit the vulnerabilities of IT infrastructure and manipulate traffic condition data.
no code implementations • 9 Aug 2022 • Yuhan Zhao, Craig Rieger, Quanyan Zhu
In this book chapter, we present a multi-agent system (MAS) framework for distributed large-scale control systems and discuss the role of MAS learning in resiliency.
1 code implementation • 29 Jul 2022 • Tao Li, Haozhe Lei, Quanyan Zhu
It leads to two online attack schemes: Intermittent Attack and Persistent Attack, which enable the attacker to learn an optimal sampling attack, defined by an $\epsilon$-first-order stationary point, within $\mathcal{O}(\epsilon^{-2})$ iterations.
no code implementations • 5 Apr 2022 • Linan Huang, Quanyan Zhu
Incentive design is a proactive and non-invasive approach to achieving compliance by aligning an insider's incentive with the defender's security objective, which motivates (rather than commands) an insider to act in the organization's interests.
no code implementations • 11 Mar 2022 • Yunhan Huang, Quanyan Zhu
The attacker can also gradually trick the ADP learner into learning the same `nefarious' policy by consistently feeding the learner a falsified cost signal that stays close to the actual cost signal.
no code implementations • 3 Mar 2022 • Yuhan Zhao, Quanyan Zhu
As on-orbit repairs are challenging, a distributed and autonomous protection mechanism is necessary to ensure the adaptation and self-healing of the satellite constellation coverage from different attacks.
no code implementations • 16 Jan 2022 • Tao Zhang, Quanyan Zhu
A revelation-principle-like design regime is established to show that the persuasion with belief hierarchies can be fully characterized by correlating the randomization of the agents' local BPD mechanisms with the persuasion as a direct recommendation of the future promises.
no code implementations • 1 Nov 2021 • Linan Huang, Quanyan Zhu
In this work, we identify and formally define a new type of proactive attentional attacks called Informational Denial-of-Service (IDoS) attacks that generate a large volume of feint attacks to overload human operators and hide real attacks among feints.
no code implementations • 4 Aug 2021 • Linan Huang, Quanyan Zhu
The numerical results illustrate how AM strategies can alleviate the severity level and the risk of IDoS attacks.
no code implementations • 2 Jul 2021 • Yunhan Huang, Linan Huang, Quanyan Zhu
In this work, we review the literature on RL for cyber resilience and discuss cyber resilience against three major types of vulnerabilities, i. e., posture-related, information-related, and human-related vulnerabilities.
no code implementations • 13 Jun 2021 • Linan Huang, Shumeng Jia, Emily Balcetis, Quanyan Zhu
The results show that the visual aids can statistically increase the attention level and improve the accuracy of phishing recognition from 74. 6% to a minimum of 86%.
no code implementations • 1 Jun 2021 • Yunhan Huang, Quanyan Zhu
In this review, we motivate the game-theoretic approach to human decision-making amid epidemics.
no code implementations • 17 May 2021 • Tao Li, Guanze Peng, Quanyan Zhu, Tamer Basar
In addition to existing research works on game-theoretic learning over networks, we highlight several new angles and research endeavors on learning in games that are related to recent developments in artificial intelligence.
no code implementations • 7 May 2021 • Tao Zhang, Quanyan Zhu
We propose a direct information design approach that incentivizes each agent to select the signal sent by the principal, such that the design process avoids the predictions of the agents' strategic selection behaviors.
no code implementations • 31 Mar 2021 • Juntao Chen, Yunhan Huang, Quanyan Zhu
Renewable energy-based microgrids play a critical role in future smart grids.
no code implementations • 24 Mar 2021 • Yunhan Huang, Juntao Chen, Quanyan Zhu
Moreover, we show that the observation choices of the defender and the attacker can be decoupled and the Nash observation strategies can be found by solving two independent optimization problems.
no code implementations • 17 Feb 2021 • Yunhan Huang, Quanyan Zhu
We study the co-design problems of the control policy and the triggering policy to optimize two pre-specified cost criteria.
no code implementations • 14 Feb 2021 • Tao Zhang, Quanyan Zhu
An obedient principle is established which states that it is without loss of generality to focus on the direct information design when the information design incentivizes each agent to select the signal sent by the designer, such that the design process avoids the predictions of the agents' strategic selection behaviors.
no code implementations • 10 Feb 2021 • Yunhan Huang, Quanyan Zhu
We also show that when the game's horizon goes to infinity, the Nash observation strategy is to observe periodically, and the expected distance between the pursuer and the evader goes to zero with a bounded second moment.
no code implementations • 4 Feb 2021 • Song Fang, Quanyan Zhu
In this paper, we relate the feedback capacity of parallel additive colored Gaussian noise (ACGN) channels to a variant of the Kalman filter.
no code implementations • 28 Dec 2020 • Tao Li, Guanze Peng, Quanyan Zhu
This work provides a novel interpretation of Markov Decision Processes (MDP) from the online optimization viewpoint.
no code implementations • 22 Dec 2020 • Song Fang, Quanyan Zhu
We first consider the scenario where the plant (i. e., the dynamical system to be controlled) is linear time-invariant, and it is seen in general that the lower bounds are characterized by the unstable poles (or nonminimum-phase zeros) of the plant as well as the conditional entropy of the disturbance.
no code implementations • 7 Dec 2020 • Song Fang, Quanyan Zhu
This short note is on a property of the $\mathcal{W}_2$ Wasserstein distance which indicates that independent elliptical distributions minimize their $\mathcal{W}_2$ Wasserstein distance from given independent elliptical distributions with the same density generators.
no code implementations • 7 Dec 2020 • Song Fang, Quanyan Zhu
In this short note, we introduce the spectral-domain $\mathcal{W}_2$ Wasserstein distance for elliptical stochastic processes in terms of their power spectra.
no code implementations • 4 Dec 2020 • Yunhan Huang, Zehui Xiong, Quanyan Zhu
On the other hand, the interactions between the attacker and the defender in the physical layer significantly impact the observation and jamming strategies.
no code implementations • 3 Dec 2020 • Song Fang, Quanyan Zhu
In this paper, we analyze the fundamental stealthiness-distortion tradeoffs of linear Gaussian dynamical systems under data injection attacks using a power spectral analysis, whereas the Kullback-Leibler (KL) divergence is employed as the stealthiness measure.
no code implementations • 29 Nov 2020 • Juntao Chen, Yunhan Huang, Rui Zhang, Quanyan Zhu
The designed curing strategy globally optimizes the trade-off between the curing cost and the severity of epidemics in the network.
no code implementations • 19 Nov 2020 • Guanze Peng, Tao Li, Shutian Liu, Juntao Chen, Quanyan Zhu
We use \textit{awareness levels} to capture the scope of the network constraints that players are aware of.
no code implementations • 4 Nov 2020 • Song Fang, Quanyan Zhu
This short note is on a property of the Kullback-Leibler (KL) divergence which indicates that independent Gaussian distributions minimize the KL divergence from given independent Gaussian distributions.
no code implementations • 29 Oct 2020 • Song Fang, Quanyan Zhu
In this paper, we study the fundamental limits of obfuscation in terms of privacy-distortion tradeoffs for linear Gaussian dynamical systems via an information-theoretic approach.
no code implementations • 12 Aug 2020 • Zhi-Li Zhang, Quanyan Zhu
This paper studies the deception applied on agent in a partially observable Markov decision process.
no code implementations • 11 Aug 2020 • Song Fang, Quanyan Zhu
In this paper, we first introduce the notion of channel leakage as the minimum mutual information between the channel input and channel output.
1 code implementation • 20 Mar 2020 • Timothy Kieras, Muhammad Junaid Farooq, Quanyan Zhu
Supply chain security threats pose new challenges to security risk modeling techniques for complex ICT systems such as the IoT.
Cryptography and Security Systems and Control Systems and Control
no code implementations • 8 Mar 2020 • Rui Zhang, Quanyan Zhu
Distributed machine learning algorithms play a significant role in processing massive data sets over large networks.
no code implementations • 7 Feb 2020 • Yunhan Huang, Quanyan Zhu
Focusing on adversarial manipulation on the cost signals, we analyze the performance degradation of TD($\lambda$) and $Q$-learning algorithms under the manipulation.
no code implementations • 12 Jan 2020 • Song Fang, Quanyan Zhu
We also investigate the implications of the results in analyzing the fundamental limits of generalization in fitting (learning) problems from the perspective of prediction with side information, as well as the fundamental limits of recursive algorithms by viewing them as generalized prediction problems.
no code implementations • 9 Jan 2020 • Song Fang, Quanyan Zhu
In this paper, we relate a feedback channel with any finite-order autoregressive moving-average (ARMA) Gaussian noises to a variant of the Kalman filter.
no code implementations • 3 Jan 2020 • Quanyan Zhu
In this chapter, we introduce methods to address resiliency issues for control systems.
no code implementations • 11 Dec 2019 • Song Fang, Quanyan Zhu
In this paper, we utilize information theory to study the fundamental performance limitations of generic feedback systems, where both the controller and the plant may be any causal functions/mappings while the disturbance can be with any distributions.
no code implementations • 6 Dec 2019 • Song Fang, Quanyan Zhu
As such, the feedback linearization together with the linear controller compose the overall relativistic feedback control law.
no code implementations • 3 Dec 2019 • Song Fang, Quanyan Zhu
In this paper, we obtain fundamental $\mathcal{L}_{p}$ bounds in sequential prediction and recursive algorithms via an entropic analysis.
1 code implementation • 28 Nov 2019 • Timothy Kieras, Muhammad Junaid Farooq, Quanyan Zhu
Securing the supply chain of information and communications technology (ICT) has recently emerged as a critical concern for national security and integrity.
Cryptography and Security
no code implementations • 11 Oct 2019 • Song Fang, Quanyan Zhu
In this paper, we derive generic bounds on the maximum deviations in prediction errors for sequential prediction via an information-theoretic approach.
no code implementations • 22 Aug 2019 • Tao Li, Quanyan Zhu
In this paper, we propose a generic framework for devising an adaptive approximation scheme for value function approximation in reinforcement learning, which introduces multiscale approximation.
no code implementations • 1 Jul 2019 • Linan Huang, Quanyan Zhu
The increasing instances of advanced attacks call for a new defense paradigm that is active, autonomous, and adaptive, named as the \texttt{`3A'} defense paradigm.
Cryptography and Security
no code implementations • 30 Jun 2019 • Linan Huang, Quanyan Zhu
The deception is rational as robots aim to achieve their deception goals at minimum cost.
no code implementations • 27 Jun 2019 • Linan Huang, Quanyan Zhu
In this work, we apply infinite-horizon Semi-Markov Decision Process (SMDP) to characterize a stochastic transition and sojourn time of attackers in the honeynet and quantify the reward-risk trade-off.
no code implementations • 24 Jun 2019 • Yunhan Huang, Quanyan Zhu
This paper studies reinforcement learning (RL) under malicious falsification on cost signals and introduces a quantitative framework of attack models to understand the vulnerabilities of RL.
no code implementations • 9 Apr 2019 • Song Fang, Mikael Skoglund, Karl Henrik Johansson, Hideaki Ishii, Quanyan Zhu
In this paper, we obtain generic bounds on the variances of estimation and prediction errors in time series analysis via an information-theoretic approach.
no code implementations • 27 Feb 2018 • Hamza Anwar, Quanyan Zhu
SVI poses variational inference as a stochastic optimization problem and solves it iteratively using noisy gradient estimates.
no code implementations • 7 Feb 2018 • Rui Zhang, Quanyan Zhu
Distributed Support Vector Machines (DSVM) have been developed to solve large-scale classification problems in networked systems with a large number of sensors and control units.
no code implementations • 12 Oct 2017 • Rui Zhang, Quanyan Zhu
The Nash equilibrium of the game allows predicting the outcome of learning algorithms in adversarial environments, and enhancing the resilience of the machine learning through dynamic distributed learning algorithms.
no code implementations • 15 Jun 2017 • Rui Zhang, Quanyan Zhu
We show that the risks of the target tasks in the nodes without the data of the source tasks can also be reduced using the information transferred from the nodes who contain the data of the source tasks.
no code implementations • 8 Jun 2017 • Jeffrey Pawlick, Quanyan Zhu
Data ecosystems are becoming larger and more complex due to online tracking, wearable computing, and the Internet of Things.
Cryptography and Security
no code implementations • 8 Aug 2016 • Jeffrey Pawlick, Quanyan Zhu
Data is the new oil; this refrain is repeated extensively in the age of internet tracking, machine learning, and data analytics.
no code implementations • 21 Jun 2016 • Jeffrey Pawlick, Quanyan Zhu
First, a machine learner declares a privacy protection level, and then users respond by choosing their own perturbation amounts.
no code implementations • 14 Jan 2016 • Tao Zhang, Quanyan Zhu
Privacy-preserving distributed machine learning becomes increasingly important due to the recent rapid growth of data.
no code implementations • 6 Jul 2013 • Wei Chen, Dayu Huang, Ankur A. Kulkarni, Jayakrishnan Unnikrishnan, Quanyan Zhu, Prashant Mehta, Sean Meyn, Adam Wierman
Neuro-dynamic programming is a class of powerful techniques for approximating the solution to dynamic programming equations.