Yinglong Ma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yinglong Ma has a total of 70 co-authors. Here are the top co-authors by number of shared publications:
- Chenqi Guo
School of Control and Computer Engineering, North China Electric Power University, Beijing, China
8 shared publications - Xiaofeng Liu
School of Control and Computer Engineering, North China Electric Power University, Beijing, China
7 shared publications - Huili Liu
School of Control and Computer Engineering, North China Electric Power University, Beijing, China
6 shared publications - Yandan Wang
North China Electric Power University, Beijing, China
5 shared publications - Tingdong Wang
North China Electric Power University
4 shared publications Beihong Jin
Institute of Software, Chinese Academy of Sciences
3 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 10799-10808
Supervision Intensity-Guided Adaptive One-to-Many Matching for Long-Tailed Object Detection
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 10799-10808
Open AccessSupervision Intensity-Guided Adaptive One-to-Many Matching for Long-Tailed Object Detection
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 17 (6) , 6071-6083
Hybrid Attention-Augmented Self-Interpretable Channel-Grouped Graph Convolutional Network for Robust Distribution Network Fault Diagnosis
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Smart Grid
- Vol. 17 (6) , 6071-6083
Hybrid Attention-Augmented Self-Interpretable Channel-Grouped Graph Convolutional Network for Robust Distribution Network Fault Diagnosis
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 41 (6) , 4252-4264
Fast and Scalable Newton-Raphson Power Flow Analysis via Hierarchical Parallelism and Graph Node-Based Hash Storage
The continuous expansion of smart grids has led to increasingly frequent power flow variations, thereby posing new demands for fast and scalable parallel implementation of Newton-Raphson power flow (N-R PF) calculations. However, existing parallel N-R PF methods often suffer from redundant floating-point operations and centralized data processing of matrix reconstruction for N-R PF calculation, which significantly increases the computational burden and impairs the parallel performance of N-R PF. To address these limitations, this paper proposes a fast and scalable hierarchical parallel strategy and graph node-based hash storage (HPGNH) method to accelerate N-R PF analysis. HPGNH integrates intra-node and inter-node parallel strategies for effectively eliminating redundant floating-point operations in Jacobian matrix updates and equation residual updates. Moreover, we present a graph node-based hash storage structure to efficiently construct and update the nonzero entries of sparse matrices in a fully parallel manner. Particularly for Q-limit N-R PF calculations, HPGNH only updates a subset of entries within the constructed storage structure instead of performing full-scale reconstruction, thereby significantly improving computational efficiency. Extensive experiments on twelve benchmark power systems demonstrate that HPGNH outperforms nine state-of-the-art methods, with computational performance improvements of 63.1% for N-R PF and 81.8% for Q-limit N-R PF.
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Power Systems
- Vol. 41 (6) , 4252-4264
Fast and Scalable Newton-Raphson Power Flow Analysis via Hierarchical Parallelism and Graph Node-Based Hash Storage
The continuous expansion of smart grids has led to increasingly frequent power flow variations, thereby posing new demands for fast and scalable parallel implementation of Newton-Raphson power flow (N-R PF) calculations. However, existing parallel N-R PF methods often suffer from redundant floating-point operations and centralized data processing of matrix reconstruction for N-R PF calculation, which significantly increases the computational burden and impairs the parallel performance of N-R PF. To address these limitations, this paper proposes a fast and scalable hierarchical parallel strategy and graph node-based hash storage (HPGNH) method to accelerate N-R PF analysis. HPGNH integrates intra-node and inter-node parallel strategies for effectively eliminating redundant floating-point operations in Jacobian matrix updates and equation residual updates. Moreover, we present a graph node-based hash storage structure to efficiently construct and update the nonzero entries of sparse matrices in a fully parallel manner. Particularly for Q-limit N-R PF calculations, HPGNH only updates a subset of entries within the constructed storage structure instead of performing full-scale reconstruction, thereby significantly improving computational efficiency. Extensive experiments on twelve benchmark power systems demonstrate that HPGNH outperforms nine state-of-the-art methods, with computational performance improvements of 63.1% for N-R PF and 81.8% for Q-limit N-R PF.
- 16 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (10) , 21005-21017
OFLT: One-Shot Federated Long-Tailed Learning With Discriminative Representation Learning and Adaptive Logit Calibration
Federated learning (FL) is a distributed machine learning algorithm that enables multiple devices to collaboratively train a global model with data privacy protection. However, in real-world scenarios, the prevalent non-independent and identically distributed (non-IID) and long-tailed data across devices results in significant performance degradation in FL. Although numerous federated long-tailed learning (Fed-LT) methods have been proposed to address this issue, they typically rely on extensive iterative communication or public datasets, resulting in high communication overhead and potential privacy and security risks. To address these challenges, we propose a one-shot Fed-LT (OFLT) framework that aims to improve the global model’s performance in FL under non-IID and long-tailed data in a single communication round. Specifically, OFLT employs class center-based contrastive learning and feature decorrelation for discriminative representation learning on clients, enhancing the intraclass compactness and interclass separability. Subsequently, supervised distillation-based generator training and adaptive logit calibration are introduced on the server to mitigate class bias in the ensemble output of local models, and the calibrated knowledge is distilled into the global model. Extensive experiments on four image datasets demonstrate that OFLT outperforms state-of-the-art (SOTA) baselines under various settings, consistently improving not only overall accuracy but also performance on medium-shot and few-shot classes.
- 16 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 13 (10) , 21005-21017
OFLT: One-Shot Federated Long-Tailed Learning With Discriminative Representation Learning and Adaptive Logit Calibration
Federated learning (FL) is a distributed machine learning algorithm that enables multiple devices to collaboratively train a global model with data privacy protection. However, in real-world scenarios, the prevalent non-independent and identically distributed (non-IID) and long-tailed data across devices results in significant performance degradation in FL. Although numerous federated long-tailed learning (Fed-LT) methods have been proposed to address this issue, they typically rely on extensive iterative communication or public datasets, resulting in high communication overhead and potential privacy and security risks. To address these challenges, we propose a one-shot Fed-LT (OFLT) framework that aims to improve the global model’s performance in FL under non-IID and long-tailed data in a single communication round. Specifically, OFLT employs class center-based contrastive learning and feature decorrelation for discriminative representation learning on clients, enhancing the intraclass compactness and interclass separability. Subsequently, supervised distillation-based generator training and adaptive logit calibration are introduced on the server to mitigate class bias in the ensemble output of local models, and the calibrated knowledge is distilled into the global model. Extensive experiments on four image datasets demonstrate that OFLT outperforms state-of-the-art (SOTA) baselines under various settings, consistently improving not only overall accuracy but also performance on medium-shot and few-shot classes.
- 4 February 2026
- Published by Elsevier in Information Sciences
- Vol. 740
Post-hoc explainability of graph neural networks: A comprehensive survey
- 4 February 2026
- Published by Elsevier in Information Sciences
- Vol. 740
Post-hoc explainability of graph neural networks: A comprehensive survey
- 15 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 22 (3) , 2853-2865
MOHFL: Multi-Level One-Shot Hierarchical Federated Learning With Enhanced Model Aggregation Over Non-IID Data
Hierarchical federated learning (HFL) is a privacy-preserving distributed machine learning framework with a client-edge-cloud hierarchy, where multiple edge servers perform partial model aggregation to reduce costly communication with the cloud server. Nevertheless, most existing HFL methods require extensive iterative communication and public datasets, which not only increase communication overhead but also raise privacy and security concerns. Moreover, non-independent and identically distributed (non-IID) data among devices can significantly impact the accuracy of the global model in HFL. To address these challenges, we propose a multi-level one-shot HFL framework (MOHFL), which aims to improve the performance of the global model in a single communication round. Specifically, we employ conditional variational autoencoders (CVAEs) as local models and use the aggregated decoders to generate an IID training set for the global model, thereby mitigating the negative impact of non-IID data. We improve the performance of CVAEs under different levels of data heterogeneity through a dominant class-based data selection method. Subsequently, an edge aggregation scheme based on multi-teacher knowledge distillation and contrastive learning is proposed to aggregate the knowledge from local decoders to edge decoders. Extensive experiments on four real-world datasets demonstrate that MOHFL is very competitive against four state-of-the-art baselines under various settings.
- 15 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 22 (3) , 2853-2865
MOHFL: Multi-Level One-Shot Hierarchical Federated Learning With Enhanced Model Aggregation Over Non-IID Data
Hierarchical federated learning (HFL) is a privacy-preserving distributed machine learning framework with a client-edge-cloud hierarchy, where multiple edge servers perform partial model aggregation to reduce costly communication with the cloud server. Nevertheless, most existing HFL methods require extensive iterative communication and public datasets, which not only increase communication overhead but also raise privacy and security concerns. Moreover, non-independent and identically distributed (non-IID) data among devices can significantly impact the accuracy of the global model in HFL. To address these challenges, we propose a multi-level one-shot HFL framework (MOHFL), which aims to improve the performance of the global model in a single communication round. Specifically, we employ conditional variational autoencoders (CVAEs) as local models and use the aggregated decoders to generate an IID training set for the global model, thereby mitigating the negative impact of non-IID data. We improve the performance of CVAEs under different levels of data heterogeneity through a dominant class-based data selection method. Subsequently, an edge aggregation scheme based on multi-teacher knowledge distillation and contrastive learning is proposed to aggregate the knowledge from local decoders to edge decoders. Extensive experiments on four real-world datasets demonstrate that MOHFL is very competitive against four state-of-the-art baselines under various settings.
- 6 April 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 279
Byzantine-robust one-shot federated learning based on hybrid-domain fuzzy clustering and meta learning
- 6 April 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 279
Byzantine-robust one-shot federated learning based on hybrid-domain fuzzy clustering and meta learning
- Open Access
Crossmodal Knowledge Distillation with WordNet-Relaxed Text Embeddings for Robust Image Classification
Crossmodal knowledge distillation (KD) aims to enhance a unimodal student using a multimodal teacher model. In particular, when the teacher's modalities include the student's, additional complementary information can be exploited to improve knowledge transfer. In supervised image classification, image datasets typically include class labels that represent high-level concepts, suggesting a natural avenue to incorporate textual cues for crossmodal KD. However, these labels rarely capture the deeper semantic structures in real-world visuals and can lead to label leakage if used directly as inputs, ultimately limiting KD performance. To address these issues, we propose a multi-teacher crossmodal KD framework that integrates CLIP image embeddings with learnable WordNet-relaxed text embeddings under a hierarchical loss. By avoiding direct use of exact class names and instead using semantically richer WordNet expansions, we mitigate label leakage and introduce more diverse textual cues. Experiments show that this strategy significantly boosts student performance, whereas noisy or overly precise text embeddings hinder distillation efficiency. Interpretability analyses confirm that WordNet-relaxed prompts encourage heavier reliance on visual features over textual shortcuts, while still effectively incorporating the newly introduced textual cues. Our method achieves state-of-the-art or second-best results on six public datasets, demonstrating its effectiveness in advancing crossmodal KD.
Open AccessCrossmodal Knowledge Distillation with WordNet-Relaxed Text Embeddings for Robust Image Classification
Crossmodal knowledge distillation (KD) aims to enhance a unimodal student using a multimodal teacher model. In particular, when the teacher's modalities include the student's, additional complementary information can be exploited to improve knowledge transfer. In supervised image classification, image datasets typically include class labels that represent high-level concepts, suggesting a natural avenue to incorporate textual cues for crossmodal KD. However, these labels rarely capture the deeper semantic structures in real-world visuals and can lead to label leakage if used directly as inputs, ultimately limiting KD performance. To address these issues, we propose a multi-teacher crossmodal KD framework that integrates CLIP image embeddings with learnable WordNet-relaxed text embeddings under a hierarchical loss. By avoiding direct use of exact class names and instead using semantically richer WordNet expansions, we mitigate label leakage and introduce more diverse textual cues. Experiments show that this strategy significantly boosts student performance, whereas noisy or overly precise text embeddings hinder distillation efficiency. Interpretability analyses confirm that WordNet-relaxed prompts encourage heavier reliance on visual features over textual shortcuts, while still effectively incorporating the newly introduced textual cues. Our method achieves state-of-the-art or second-best results on six public datasets, demonstrating its effectiveness in advancing crossmodal KD.