Zhongyan Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Zhongyan Li has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
- Han Su
School of Control and Computer Engineering, North China Electric Power University, Beijing, China
3 shared publications - Huixian Meng
North China Electric Power University
3 shared publications - Wanquan Liu
School of Intelligent Systems Engineering, Shenzhen Campus of Sun Yat-sen University, Shenzhen, China
3 shared publications - Xian-Liang Shi
Hu’nan Normal University
3 shared publications - Yuanan Diao
Department of Mathematics and Statistics, University of North Carolina Charlotte, Charlotte, NC 28223, USA
3 shared publications - Deguang Han
University of Central Florida
2 shared publications
Latest Publications
- 7 October 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (40) , 405102
Robust steerability classification via key feature extraction
Generalization is essential for reliable machine-learning detection of quantum steerability. In this work, we develop a CNN-based approach using the matrix versions of the full-information feature F 1 and the steering-determined feature F 2 as inputs for two-qubit steerability classification. Several physically distinct state families with reliable labels are employed to evaluate classifier generalization. Vector-input classifiers based on F 1 and F 2 are included as benchmarks under a unified classification protocol. Across the considered classifiers and test datasets, F 2 -based models exhibit consistently more stable generalization than their F 1 -based counterparts, indicating that steering-determined feature extraction plays a more dominant role than classifier selection in the observed performance improvement in the considered settings. Meanwhile, no single classifier is uniformly optimal across all physical state families. Since CNN– F 2 achieves the best overall performance on the T -diagonal benchmarks, it is further employed to visualize and estimate the projective-measurement steering hierarchy of axially symmetric states, a special subclass of T -diagonal states.
- 7 October 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (40) , 405102
Robust steerability classification via key feature extraction
Generalization is essential for reliable machine-learning detection of quantum steerability. In this work, we develop a CNN-based approach using the matrix versions of the full-information feature F 1 and the steering-determined feature F 2 as inputs for two-qubit steerability classification. Several physically distinct state families with reliable labels are employed to evaluate classifier generalization. Vector-input classifiers based on F 1 and F 2 are included as benchmarks under a unified classification protocol. Across the considered classifiers and test datasets, F 2 -based models exhibit consistently more stable generalization than their F 1 -based counterparts, indicating that steering-determined feature extraction plays a more dominant role than classifier selection in the observed performance improvement in the considered settings. Meanwhile, no single classifier is uniformly optimal across all physical state families. Since CNN– F 2 achieves the best overall performance on the T -diagonal benchmarks, it is further employed to visualize and estimate the projective-measurement steering hierarchy of axially symmetric states, a special subclass of T -diagonal states.
- Open Access
- 10 February 2026
- Published by Elsevier in SSRN Electronic Journal
Keyword-Speaker-Extraction-Generation (KSEG): A two-stage summarizing framework for Chinese online medical dialogue text
The exponential growth of online medical consultations has generated millions of dialogues, yet automated summarization faces challenges from conversational noise, sparse key information, and stringent terminological precision. Existing methods often fail in fine-grained noise filtering, faithful medical entity reproduction, and computational efficiency. To address these issues, we propose KSEG (Keyword-Speaker-Extraction-Generation), a two-stage neural framework. The extraction stage employs fine-grained segment-level processing and medical term density weighting, utilizing Robustly Optimized Bidirectional Encoder Representation from Transformers Pretraining Approach (RoBERTa) with Dilated Gated Convolutional Networks (DGCNN) to reduce input length by 70%. The generation stage combines a Neural Contextualized Representation for Chinese Language Understanding (NEZHA) encoder with a Unified Language Model (UniLM) decoder, augmented by a novel BIO-COPY mechanism. This mechanism uses structured tri-label control (generate/copy-start/copy-continue) to explicitly ensure the accurate reproduction of medical entities. Evaluated on the IMCS-21 dataset, KSEG achieved Recall-Oriented Understudy for Gisting Evaluation (ROUGE-1/2/L) scores of 61.37/47.33/57.66, outperforming strong baselines by 3-11 points and matching some large language models with only 200 million parameters. Ablation studies confirmed the BIO-COPY mechanism's critical role, with its removal causing a 7.59-point ROUGE-1 drop. Human evaluation also confirmed its superior factual accuracy and reduced hallucinations, establishing a resource-efficient paradigm for clinical settings.
- 10 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessKeyword-Speaker-Extraction-Generation (KSEG): A two-stage summarizing framework for Chinese online medical dialogue text
The exponential growth of online medical consultations has generated millions of dialogues, yet automated summarization faces challenges from conversational noise, sparse key information, and stringent terminological precision. Existing methods often fail in fine-grained noise filtering, faithful medical entity reproduction, and computational efficiency. To address these issues, we propose KSEG (Keyword-Speaker-Extraction-Generation), a two-stage neural framework. The extraction stage employs fine-grained segment-level processing and medical term density weighting, utilizing Robustly Optimized Bidirectional Encoder Representation from Transformers Pretraining Approach (RoBERTa) with Dilated Gated Convolutional Networks (DGCNN) to reduce input length by 70%. The generation stage combines a Neural Contextualized Representation for Chinese Language Understanding (NEZHA) encoder with a Unified Language Model (UniLM) decoder, augmented by a novel BIO-COPY mechanism. This mechanism uses structured tri-label control (generate/copy-start/copy-continue) to explicitly ensure the accurate reproduction of medical entities. Evaluated on the IMCS-21 dataset, KSEG achieved Recall-Oriented Understudy for Gisting Evaluation (ROUGE-1/2/L) scores of 61.37/47.33/57.66, outperforming strong baselines by 3-11 points and matching some large language models with only 200 million parameters. Ablation studies confirmed the BIO-COPY mechanism's critical role, with its removal causing a 7.59-point ROUGE-1 drop. Human evaluation also confirmed its superior factual accuracy and reduced hallucinations, establishing a resource-efficient paradigm for clinical settings.
- Open Access
The Color Image Watermarking Algorithm Based on Quantum Discrete Wavelet Transform and Chaotic Mapping
Quantum watermarking is a technique that embeds specific information into a quantum carrier for the purpose of digital copyright protection. In this paper, we propose a novel color image watermarking algorithm that integrates quantum discrete wavelet transform with Sinusoidal–Tent mapping and baker mapping. Initially, chaotic sequences are generated using Sinusoidal–Tent mapping to determine the channels suitable for watermark embedding. Subsequently, a one-level quantum Haar wavelet transform is applied to the selected channel to decompose the image. The watermarked image is then scrambled via discrete baker mapping, and the scrambled image is embedded into the High-High subbands. The invisibility of the watermark is evaluated by calculating the peak signal-to-noise ratio, Structural similarity index measure, and Learned Perceptual Image Patch Similarity, with comparisons made against the color histogram. The robustness of the proposed algorithm is assessed through the calculation of Normalized Cross-Correlation. In the simulation results, PSNR is close to 63, SSIM is close to 1, LPIPS is close to 0.001, and NCC is close to 0.97. This indicates that the proposed watermarking algorithm exhibits excellent visual quality and a robust capability to withstand various attacks. Additionally, through ablation study, the contribution of each technique to overall performance was systematically evaluated.
Open AccessThe Color Image Watermarking Algorithm Based on Quantum Discrete Wavelet Transform and Chaotic Mapping
Quantum watermarking is a technique that embeds specific information into a quantum carrier for the purpose of digital copyright protection. In this paper, we propose a novel color image watermarking algorithm that integrates quantum discrete wavelet transform with Sinusoidal–Tent mapping and baker mapping. Initially, chaotic sequences are generated using Sinusoidal–Tent mapping to determine the channels suitable for watermark embedding. Subsequently, a one-level quantum Haar wavelet transform is applied to the selected channel to decompose the image. The watermarked image is then scrambled via discrete baker mapping, and the scrambled image is embedded into the High-High subbands. The invisibility of the watermark is evaluated by calculating the peak signal-to-noise ratio, Structural similarity index measure, and Learned Perceptual Image Patch Similarity, with comparisons made against the color histogram. The robustness of the proposed algorithm is assessed through the calculation of Normalized Cross-Correlation. In the simulation results, PSNR is close to 63, SSIM is close to 1, LPIPS is close to 0.001, and NCC is close to 0.97. This indicates that the proposed watermarking algorithm exhibits excellent visual quality and a robust capability to withstand various attacks. Additionally, through ablation study, the contribution of each technique to overall performance was systematically evaluated.
- Open Access
Broad stochastic configuration residual learning system for norm-convergent universal approximation
Universal approximation serves as the foundation of neural network learning algorithms. However, some networks establish their universal approximation property by demonstrating that the iterative errors converge in probability measure rather than the more rigorous norm convergence, which makes the universal approximation property of randomized learning networks highly sensitive to random parameter selection, Broad residual learning system (BRLS), as a member of randomized learning models, also encounters this issue. We theoretically demonstrate the limitation of its universal approximation property, that is, the iterative errors do not satisfy norm convergence if the selection of random parameters is inappropriate and the convergence rate meets certain conditions. To address this issue, we propose the broad stochastic configuration residual learning system (BSCRLS) algorithm, which features a novel supervisory mechanism adaptively constraining the range settings of random parameters on the basis of BRLS framework, Furthermore, we prove the universal approximation theorem of BSCRLS based on the more stringent norm convergence. Three versions of incremental BSCRLS algorithms are presented to satisfy the application requirements of various network updates. Solar panels dust detection experiments are performed on publicly available dataset and compared with 13 deep and broad learning algorithms. Experimental results reveal the effectiveness and superiority of BSCRLS algorithms.
Open AccessBroad stochastic configuration residual learning system for norm-convergent universal approximation
Universal approximation serves as the foundation of neural network learning algorithms. However, some networks establish their universal approximation property by demonstrating that the iterative errors converge in probability measure rather than the more rigorous norm convergence, which makes the universal approximation property of randomized learning networks highly sensitive to random parameter selection, Broad residual learning system (BRLS), as a member of randomized learning models, also encounters this issue. We theoretically demonstrate the limitation of its universal approximation property, that is, the iterative errors do not satisfy norm convergence if the selection of random parameters is inappropriate and the convergence rate meets certain conditions. To address this issue, we propose the broad stochastic configuration residual learning system (BSCRLS) algorithm, which features a novel supervisory mechanism adaptively constraining the range settings of random parameters on the basis of BRLS framework, Furthermore, we prove the universal approximation theorem of BSCRLS based on the more stringent norm convergence. Three versions of incremental BSCRLS algorithms are presented to satisfy the application requirements of various network updates. Solar panels dust detection experiments are performed on publicly available dataset and compared with 13 deep and broad learning algorithms. Experimental results reveal the effectiveness and superiority of BSCRLS algorithms.
- 11 October 2025
- Published by Elsevier in Renewable and Sustainable Energy Reviews
- Vol. 226
A hybrid prediction model of photovoltaic power system based on AP, ISSA-based VMD, CLKAN and error correction
- 11 October 2025
- Published by Elsevier in Renewable and Sustainable Energy Reviews
- Vol. 226
A hybrid prediction model of photovoltaic power system based on AP, ISSA-based VMD, CLKAN and error correction
- 9 October 2025
- Published by American Physical Society (APS) in Physical Review A
- Vol. 112 (4)
Machine learning for detecting steering in qutrit-pair states
- 9 October 2025
- Published by American Physical Society (APS) in Physical Review A
- Vol. 112 (4)
Machine learning for detecting steering in qutrit-pair states
- 15 July 2025
- Published by Elsevier in Pattern Recognition
- Vol. 171
A novel broad learning network with residual technique
- 15 July 2025
- Published by Elsevier in Pattern Recognition
- Vol. 171
A novel broad learning network with residual technique