Source author record

Yuzhe Li

Yuzhe Li appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Systems and Control math.OC Artificial Intelligence Cryptography and Security eess.IV eess.SY Machine Learning physics.optics

Catalog footprint

What is connected

6works

8topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Towards Practical Differential Privacy in Data Analysis: Understanding the Effect of Epsilon on Utility in Private ERM

In this paper, we focus our attention on private Empirical Risk Minimization (ERM), which is one of the most commonly used data analysis method. We take the first step towards solving the above problem by theoretically exploring the effect of epsilon (the parameter of differential privacy that determines the strength of privacy guarantee) on utility of the learning model. We trace the change of utility with modification of epsilon and reveal an established relationship between epsilon and utility. We then formalize this relationship and propose a practical approach for estimating the utility under an arbitrary value of epsilon. Both theoretical analysis and experimental results demonstrate high estimation accuracy and broad applicability of our approach in practical applications. As providing algorithms with strong utility guarantees that also give privacy when possible becomes more and more accepted, our approach would have high practical value and may be likely to be adopted by companies and organizations that would like to preserve privacy but are unwilling to compromise on utility.

preprint2021arXiv

Optimal Unbiased Linear Sensor Fusion over Multiple Lossy Channels with Collective Observability

In this paper, we consider optimal linear sensor fusion for obtaining a remote state estimate of a linear process based on the sensor data transmitted over lossy channels. There is no local observability guarantee for any of the sensors. It is assumed that the state of the linear process is collectively observable. We transform the problem of finding the optimal linear sensor fusion coefficients as a convex optimization problem which can be efficiently solved. Moreover, the closed-form expression is also derived for the optimal coefficients. Simulation results are presented to illustrate the performance of the developed algorithm.

preprint2020arXiv

Displacement-agnostic coherent imaging through scatter with an interpretable deep neural network

Coherent imaging through scatter is a challenging task in computational imaging. Both model-based and data-driven approaches have been explored to solve the inverse scattering problem. In our previous work, we have shown that a deep learning approach can make high-quality and highly generalizable predictions through unseen diffusers. Here, we propose a new deep neural network (DNN) model that is agnostic to a broader class of perturbations including scatterer change, displacements, and system defocus up to 10X depth of field. In addition, we develop a new analysis framework for interpreting the mechanism of our DNN model and visualizing its generalizability based on an unsupervised dimension reduction technique. We show that our DNN can unmix the scattering-specific information and extract the object-specific information so as to achieve generalization under different scattering conditions. Our work paves the way to a highly robust and interpretable deep learning approach to imaging through scattering media.

preprint2015arXiv

Data-Driven Power Control for State Estimation: A Bayesian Inference Approach

We consider sensor transmission power control for state estimation, using a Bayesian inference approach. A sensor node sends its local state estimate to a remote estimator over an unreliable wireless communication channel with random data packet drops. As related to packet dropout rate, transmission power is chosen by the sensor based on the relative importance of the local state estimate. The proposed power controller is proved to preserve Gaussianity of local estimate innovation, which enables us to obtain a closed-form solution of the expected state estimation error covariance. Comparisons with alternative non data-driven controllers demonstrate performance improvement using our approach.

preprint2015arXiv

Fake-Acknowledgment Attack on ACK-based Sensor Power Schedule for Remote State Estimation

We consider a class of malicious attacks against remote state estimation. A sensor with limited resources adopts an acknowledgement (ACK)-based online power schedule to improve the remote state estimation performance. A malicious attacker can modify the ACKs from the remote estimator and convey fake information to the sensor. When the capability of the attacker is limited, we propose an attack strategy for the attacker and analyze the corresponding effect on the estimation performance. The possible responses of the sensor are studied and a condition for the sensor to discard ACKs and switch from online schedule to offline schedule is provided.

preprint2014arXiv

Transmission Power Scheduling for Energy Harvesting Sensor in Remote State Estimation

We study remote estimation in a wireless sensor network. Instead of using a conventional battery-powered sensor, a sensor equipped with an energy harvester which can obtain energy from the external environment is utilized. We formulate this problem into an infinite time-horizon Markov decision process and provide the optimal sensor transmission power control strategy. In addition, a sub-optimal strategy which is easier to implement and requires less computation is presented. A numerical example is provided to illustrate the implementation of the sub-optimal policy and evaluation of its estimation performance.