About

Mingxia Gu is an Associate Professor in the Department of Anesthesiology & Perioperative Medicine at UCLA, affiliated with the Broad Stem Cell Research Center. Her research focuses on heart and lung diseases, stem cell biology, and related biomedical fields. She holds an MD and PhD from Peking University in Beijing, China, with joint training at Stanford University completed in 2013. She also completed a fellowship at Stanford University in 2016 and an instructorship there in 2019. Her work involves investigating cellular and molecular mechanisms underlying cardiovascular and pulmonary conditions, contributing to advancements in regenerative medicine and stem cell therapies.

Research topics

• Computer Science
• Data Mining
• Artificial Intelligence
• Algorithm
• Machine Learning
• Mathematics
• Combinatorics
• Theoretical computer science
• Pure mathematics

Selected publications

• Towards Scalable Web Accessibility Audit with MLLMs as Copilots

  Proceedings of the AAAI Conference on Artificial Intelligence · 2026-03-14 · 1 citations

  articleOpen access1st authorCorresponding

  Ensuring web accessibility is crucial for advancing social welfare, justice, and equality in digital spaces, yet the vast majority of website user interfaces remain non-compliant, due in part to the resource-intensive and unscalable nature of current auditing practices. While WCAG-EM offers a structured methodology for site-wise conformance evaluation, it involves great human efforts and lacks practical support for execution at scale. In this work, we present an auditing framework, AAA, which operationalizes WCAG-EM through a human-AI partnership model. AAA is anchored by two key innovations: GRASP, a graph-based multimodal sampling method that ensures representative page coverage via learned embeddings of visual, textual, and relational cues; and MaC, a multimodal large language model-based copilot strategy that supports auditors through cross-modal reasoning and intelligent assistance in high-effort tasks. Together, these components enable scalable, end-to-end web accessibility auditing, empowering human auditors with AI-enhanced assistance for real-world impact. We further contribute four novel datasets designed for benchmarking core stages of the audit pipeline. Extensive experiments demonstrate the effectiveness of our methods, providing insights that small-scale language models can serve as capable experts when fine-tuned.

  PublisherOA PDFDOI

• Improved Kalman Filtering and Adaptive Weighted Fusion Algorithms for Enhanced Multi-Sensor Data Fusion in Precision Measurement

  Informatica · 2025-01-28 · 6 citations

  articleOpen access1st authorCorresponding

  Multi-sensor data fusion plays a crucial role in achieving accurate and reliable measurements in precision measurement systems. This study focuses on the application of multi-source data fusion technology based on an improved Kalman filtering algorithm in precision measurement. The fundamental principles and structural models of multi-sensor data fusion are analyzed, highlighting the importance of effective fusion algorithms. Improvements are proposed for the weighted information fusion algorithm and the Kalman filtering fusion algorithm to enhance their performance in handling uncertainties and inconsistencies in sensor data. The improved weighted information fusion algorithm combines the Jackknife method with an adaptive weighting approach, while the improved Kalman filtering fusion algorithm incorporates a weight factor, a state transition matrix, a measurement transition matrix, and a process noise distribution matrix. The effectiveness of the improved algorithms is validated through simulations and practical applications, demonstrating significant improvements in estimation accuracy, precision, and robustness compared to traditional methods. The study also discusses the challenges and opportunities for further research in multi-sensor data fusion, including scalability, computational efficiency, and the integration of advanced techniques such as machine learning and deep learning. The findings contribute to the advancement of multi-sensor data fusion techniques and their applications in precision measurement, providing insights for future research and development.

  PublisherOA PDFDOI

• Towards an Inclusive Mobile Web: A Dataset and Framework for Focusability in UI Accessibility

  2025-04-22 · 1 citations

  article1st authorCorresponding

  The rapid growth of mobile web technologies has revolutionized how people manage daily activities, emphasizing the critical need for accessible mobile user interfaces (UIs) that accommodate users with disabilities and situational impairments. Current AI-driven UI understanding methods show promise but primarily target general UI modeling, neglecting nuanced, user-centric accessibility requirements. To bridge this gap, we first conducted a formative study with 12 visually impaired participants. Our study uncovers selective-accessible issues, a new class of accessibility challenges requiring finer granularity and selective focus on UI components, which existing methods largely overlook. Our findings also reveal that the severity of issues varies across interaction stages, with earlier stages posing a more significant impact. Building on these insights, we propose a comprehensive framework of three accessibility stages: focusability, information, and functionality (FIF), encompassing 12 sub-tasks under 3 overarching tasks. Identifying UI element focusability prediction (UFP) as a pivotal yet underexplored task within FIF, hindered by the absence of dedicated datasets, we introduce a new dataset (NOS) with 117,480 annotated components addressing accessibility issues comprehensively. To further enhance UFP, we introduce Graph-based UI Focusability Prediction (GIFT), a method leveraging graph neural networks to model UFP-targeted UI relationships. User studies validate the dataset's quality, while experiments show GIFT's effectiveness in improving UFP outcomes. Our code and datasets are publicly available to support further web inclusivity advancements at https://github.com/eaglelab-zju/NOS.

  PublisherDOI

• Frequency Self-Adaptation Graph Neural Network for Unsupervised Graph Anomaly Detection

  Neural Networks · 2025-05-27 · 3 citations

  article1st author
  PublisherDOI

• Heterophilous distribution propagation for Graph Neural Networks

  Neural Networks · 2024-12-24 · 3 citations

  article
  PublisherDOI

• Homophily-enhanced Structure Learning for Graph Clustering

  2023 · 24 citations

  1st authorCorresponding
  • Computer Science
  • Computer Science
  • Machine Learning

  Graph clustering is a fundamental task in graph analysis, and recent advances in utilizing graph neural networks (GNNs) have shown impressive results. Despite the success of existing GNN-based graph clustering methods, they often overlook the quality of graph structure, which is inherent in real-world graphs due to their sparse and multifarious nature, leading to subpar performance. Graph structure learning allows refining the input graph by adding missing links and removing spurious connections. However, previous endeavors in graph structure learning have predominantly centered around supervised settings, and cannot be directly applied to our specific clustering tasks due to the absence of ground-truth labels. To bridge the gap, we propose a novel method called homophily-enhanced structure learning for graph clustering (HoLe). Our motivation stems from the observation that subtly enhancing the degree of homophily within the graph structure can significantly improve GNNs and clustering outcomes. To realize this objective, we develop two clustering-oriented structure learning modules, i.e., hierarchical correlation estimation and cluster-aware sparsification. The former module enables a more accurate estimation of pairwise node relationships by leveraging guidance from latent and clustering spaces, while the latter one generates a sparsified structure based on the similarity matrix and clustering assignments. Additionally, we devise a joint optimization approach alternating between training the homophily-enhanced structure learning and GNN-based clustering, thereby enforcing their reciprocal effects. Extensive experiments on seven benchmark datasets of various types and scales, across a range of clustering metrics, demonstrate the superiority of HoLe against state-of-the-art baselines.

  PublisherDOI

• An Efficient and Reliable Tolerance- Based Algorithm for Principal Component Analysis

  2022 IEEE International Conference on Data Mining Workshops (ICDMW) · 2022 · 6 citations

  Senior authorCorresponding
  • Computer Science
  • Computer Science
  • Algorithm

  Principal component analysis (PCA) is an important method for dimensionality reduction in data science and machine learning. However, it is expensive for large matrices when only a few components are needed. Existing fast PCA algorithms typically assume the user will supply the number of components needed, but in practice, they may not know this number beforehand. Thus, it is important to have fast PCA algorithms depending on a tolerance. We develop one such algorithm that runs quickly for matrices with rapidly decaying singular values, provide approximation error bounds that are within a constant factor away from optimal, and demonstrate its utility with data from a variety of applications.

  PublisherDOI

• Randomized Projection for Rank-Revealing Matrix Factorizations and Low-Rank Approximations

  arXiv (Cornell University) · 2020-08-10

  preprintOpen accessSenior author

  Rank-revealing matrix decompositions provide an essential tool in spectral analysis of matrices, including the Singular Value Decomposition (SVD) and related low-rank approximation techniques. QR with Column Pivoting (QRCP) is usually suitable for these purposes, but it can be much slower than the unpivoted QR algorithm. For large matrices, the difference in performance is due to increased communication between the processor and slow memory, which QRCP needs in order to choose pivots during decomposition. Our main algorithm, Randomized QR with Column Pivoting (RQRCP), uses randomized projection to make pivot decisions from a much smaller sample matrix, which we can construct to reside in a faster level of memory than the original matrix. This technique may be understood as trading vastly reduced communication for a controlled increase in uncertainty during the decision process. For rank-revealing purposes, the selection mechanism in RQRCP produces results that are the same quality as the standard algorithm, but with performance near that of unpivoted QR (often an order of magnitude faster for large matrices). We also propose two formulas that facilitate further performance improvements. The first efficiently updates sample matrices to avoid computing new randomized projections. The second avoids large trailing updates during the decomposition in truncated low-rank approximations. Our truncated version of RQRCP also provides a key initial step in our truncated SVD approximation, TUXV. These advances open up a new performance domain for large matrix factorizations that will support efficient problem-solving techniques for challenging applications in science, engineering, and data analysis.

  PublisherOA PDFDOI

• Randomized Projection for Rank-Revealing Matrix Factorizations and Low-Rank Approximations

  SIAM Review · 2020 · 19 citations

  Senior authorCorresponding
  • Computer Science
  • Algorithm
  • Computer Science

  Rank-revealing matrix decompositions provide an essential tool in spectral analysis of matrices, including the Singular Value Decomposition (SVD) and related low-rank approximation techniques. QR with Column Pivoting (QRCP) is usually suitable for these purposes, but it can be much slower than the unpivoted QR algorithm. For large matrices, the difference in performance is due to increased communication between the processor and slow memory, which QRCP needs in order to choose pivots during decomposition. Our main algorithm, Randomized QR with Column Pivoting (RQRCP), uses randomized projection to make pivot decisions from a much smaller sample matrix, which we can construct to reside in a faster level of memory than the original matrix. This technique may be understood as trading vastly reduced communication for a controlled increase in uncertainty during the decision process. For rank-revealing purposes, the selection mechanism in RQRCP produces results that are the same quality as the standard algorithm, but with performance near that of unpivoted QR (often an order of magnitude faster for large matrices). We also propose two formulas that facilitate further performance improvements. The first efficiently updates sample matrices to avoid computing new randomized projections. The second avoids large trailing updates during the decomposition in truncated low-rank approximations. Our truncated version of RQRCP also provides a key initial step in our truncated SVD approximation, TUXV. These advances open up a new performance domain for large matrix factorizations that will support efficient problem-solving techniques for challenging applications in science, engineering, and data analysis.

  PublisherOA PDFDOI

• Efficient Spectrum-Revealing CUR Matrix Decomposition.

  International Conference on Artificial Intelligence and Statistics · 2020-06-03 · 2 citations

  article
  Publisher

Recent grants

• Fast Numerically Stable Matrix Algorithms

  NSF · $444k · 2002–2006

• Collaborative Research: Minimum Sobolov Norm Methods

  NSF · $308k · 2008–2013

• Collaborative Research: Super-fast Direct Sparse Solvers

  NSF · $125k · 2005–2009

• "AF:Small:Efficient and reliable low-rank approximation techniques and fast solutions to large sparse linear equations"

  NSF · $400k · 2013–2017

Frequent coauthors

• Stanley C. Eisenstat

  22 shared

• Shivkumar Chandrasekaran

  15 shared

• James Demmel

  13 shared

• Jianlin Xia

  10 shared

• Horst D. Simon

  Adaptive Dynamics (United States)

  10 shared

• S. Chandrasekaran

  9 shared

• Xuebin Chi

  Chinese Academy of Sciences

  9 shared

• Ali H. Sayed

  École Polytechnique Fédérale de Lausanne

  9 shared

Education

• Ph.D., Electrical Engineering

  University of California, Los Angeles

  2000

• M.S., Electrical Engineering

  University of California, Los Angeles

  1996

• B.S., Electrical Engineering

  University of Science and Technology of China

  1993