skip to main content


Search for: All records

Creators/Authors contains: "Zhang, Xiang"

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Free, publicly-accessible full text available May 1, 2025
  2. Free, publicly-accessible full text available December 27, 2024
  3. Uncovering rationales behind predictions of graph neural networks (GNNs) has received increasing attention over recent years. Instance-level GNN explanation aims to discover critical input elements, such as nodes or edges, that the target GNN relies upon for making predictions. Though various algorithms are proposed, most of them formalize this task by searching the minimal subgraph, which can preserve original predictions. However, an inductive bias is deep-rooted in this framework: Several subgraphs can result in the same or similar outputs as the original graphs. Consequently, they have the danger of providing spurious explanations and failing to provide consistent explanations. Applying them to explain weakly performed GNNs would further amplify these issues. To address this problem, we theoretically examine the predictions of GNNs from the causality perspective. Two typical reasons for spurious explanations are identified: confounding effect of latent variables like distribution shift and causal factors distinct from the original input. Observing that both confounding effects and diverse causal rationales are encoded in internal representations,we propose a new explanation framework with an auxiliary alignment loss, which is theoretically proven to be optimizing a more faithful explanation objective intrinsically. Concretely for this alignment loss, a set of different perspectives are explored: anchor-based alignment, distributional alignment based on Gaussian mixture models, mutual-information-based alignment, and so on. A comprehensive study is conducted both on the effectiveness of this new framework in terms of explanation faithfulness/consistency and on the advantages of these variants. For our codes, please refer to the following URL link:https://github.com/TianxiangZhao/GraphNNExplanation

     
    more » « less
    Free, publicly-accessible full text available October 31, 2024
  4. Free, publicly-accessible full text available August 4, 2024
  5. Free, publicly-accessible full text available April 30, 2024
  6. We consider the problem of spectrum sharing by multiple cellular operators. We propose a novel deep Reinforcement Learning (DRL)-based distributed power allocation scheme which utilizes the multi-agent Deep Deterministic Policy Gradient (MA-DDPG) algorithm. In particular, we model the base stations (BSs) that belong to the multiple operators sharing the same band, as DRL agents that simultaneously determine the transmit powers to their scheduled user equipment (UE) in a synchronized manner. The power decision of each BS is based on its own observation of the radio environment (RF) environment, which consists of interference measurements reported from the UEs it serves, and a limited amount of information obtained from other BSs. One advantage of the proposed scheme is that it addresses the single-agent non-stationarity problem of RL in the multi-agent scenario by incorporating the actions and observations of other BSs into each BS's own critic which helps it to gain a more accurate perception of the overall RF environment. A centralized-training-distributed-execution framework is used to train the policies where the critics are trained over the joint actions and observations of all BSs while the actor of each BS only takes the local observation as input in order to produce the transmit power. Simulation with the 6 GHz Unlicensed National Information Infrastructure (U-NII)-5 band shows that the proposed power allocation scheme can achieve better throughput performance than several state-of-the-art approaches. 
    more » « less
  7. In this manuscript, we present a multiscale Adaptive Reduced-Order Modeling (AROM) framework to efficiently simulate the response of heterogeneous composite microstructures under interfacial and volumetric damage. This framework builds on the eigendeformation-based reduced-order homogenization model (EHM), which is based on the transformation field analysis (TFA) and operates in the context of computational homogenization with a focus on model order reduction of the microscale problem. EHM pre-computes certain microstructure information by solving a series of linear elastic problems defined over the fully resolved microstructure (i.e., concentration tensors, interaction tensors) and approximates the microscale problem using a much smaller basis spanned over subdomains (also called parts) of the microstructure. Using this reduced basis, and prescribed spatial variation of inelastic response fields over the parts, the microscale problem leads to a set of algebraic equations with part-wise responses as unknowns, instead of node-wise displacements as in finite element analysis. The volumetric and interfacial influence functions are calculated by using the Interface enriched Generalized Finite Element Method (IGFEM) to compute the coefficient tensors, in which the finite element discretization does not need to conform to the material interfaces. AROM takes advantage of pre-computed coefficient tensors associated with the finest ROM and efficiently computes the coefficient tensors of a series of gradually coarsening ROMs. During the multiscale analysis stage, the simulation starts with a coarse ROM which can capture the initial elastic response well. As the loading continues and response in certain parts of the microstructure starts to localize, the analysis adaptively switches to the next level of refined ROM to better capture those local responses. The performance of AROM is evaluated by comparing the results with regular EHM (no adaptive refinement) and IGFEM under different loading conditions and failure modes for various 2D and 3D microstructures. The proposed AROM provides an efficient way to model history-dependent nonlinear responses for composite materials under localized interface failure and phase damage. 
    more » « less