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


Yumin Dong has a total of 115 co-authors. Here are the top co-authors by number of shared publications:

  • Dingkang Mou

    College of Computer and Information Science, Chongqing Normal University, People's Republic of China

    8 shared publications
  • Jinlei Zhang

    College of Computer and Information Science, Chongqing Normal University, Chongqing, China

    7 shared publications
  • Mingqiu Wu

    College of Computer and Information Science Chongqing Normal University Chongqing China

    6 shared publications
  • Xuanxuan Che

    College of Computer and Information Sciences, Chongqing Normal University, Chongqing, People's Republic of China

    6 shared publications
  • Rui Yan

    College of Computer and Information Science, Chongqing Normal University, Chongqing, China

    5 shared publications
  • Yanying Fu

    College of Computer and Information Science, Chongqing Normal University, Chongqing, China

    5 shared publications

Latest Publications

  • A Color Image Encryption Scheme Based on a Four-Dimensional Discrete Chaotic System with Dynamic Channel Permutation and Bidirectional Cross-Diffusion

    by Yunan Yang, Yumin Dong

    Color image encryption must simultaneously suppress spatial redundancy, adjacent-pixel correlation, and inter-channel dependence. However, many existing schemes based on independent RGB processing, simple cross-channel permutation, and unidirectional diffusion still provide limited spatial–channel collaborative mixing. To address this issue, this paper proposes a color image encryption scheme driven by a four-dimensional discrete chaotic system. The proposed design integrates a four-dimensional chaotic source with sinusoidal modulation and feedback coupling, plaintext-related global scrambling with 24-mode dynamic channel mixing, and bidirectional cross-diffusion within a unified collaborative framework. In this framework, the chaotic system provides complex control sequences, the scrambling and channel-mixing stages jointly reorganize pixel positions and RGB relationships, and the bidirectional diffusion stage strengthens perturbation propagation across both spatial positions and color channels. Dynamical analyses, NIST SP800-22 tests, and the 0–1 test verify the complexity and randomness of the chaotic source, with all median K values higher than 0.998. Experimental results show that the average ciphertext entropy, NPCR, and UACI reach 7.999, 99.655%, and 33.442%, respectively. These results indicate that the proposed collaborative framework achieves effective statistical concealment, differential sensitivity, and spatial–channel decorrelation without excessive structural complexity.

  • An enhanced classification framework based on quantum resonance dimensionality reduction and hybrid quantum ensemble learning

    by Gaojie Wu, Yumin Dong

    With the rapid growth of data volume and complexity, traditional machine learning algorithms encounter significant challenges in handling large-scale and high-dimensional datasets. To address these limitations, this paper proposes an enhanced classification framework based on Quantum Resonance Dimensionality Reduction (QRDR) and Hybrid Quantum Ensemble Learning (HQEL). First, a Multi-Kernel Quantum Support Vector Machine (MKQSVM) is introduced, which integrates the advantages of multiple quantum kernel functions and employs Bayesian optimization to automatically determine the optimal kernel combination. This approach enhances the model’s adaptability and generalization ability on heterogeneous data. Second, the hybrid quantum ensemble learning model combines MKQSVM and Quantum Convolutional Neural Network (QCNN) through a stacking ensemble strategy with a classical meta-learner, thereby achieving improved prediction accuracy and stability. Moreover, the proposed QRDR algorithm is utilized to reduce the dimensionality of high-dimensional data while retaining critical features and minimizing redundancy, which effectively improves both training efficiency and predictive performance of the quantum models. Experimental results demonstrate that the proposed method outperforms traditional machine learning approaches on multiple benchmark datasets, particularly in classifying complex and high-dimensional data.

  • FMCNS: Flow Matching with Causal-aware Negative Sampling for Multimodal Recommendation

    by Xinji Zha, Yongfu Zha, Haomiao Jiang, Xueliang Song, Huanxin Ding, Yumin Dong

  • Content Caching for IoT Devices by Using Self-Feedback Adversarial Semi-Bandits Learning

    by Jinpeng Li, Peng Chen, Yunni Xia, MengChu Zhou, Yong Ma, Yumin Dong, Hui Liu, Qinglan Peng, Xifeng Xu

    As massive data is generated by Internet of Things (IoT) devices, user-end devices are required to implement computation-intensive functionalities, including multi-sensory data processing and analysis, sophisticated system control schemes, and artificial intelligence. Mobile Edge Computing (MEC) is a significant technology that has the potential to extend the computation and storage capacities of user-end devices by the decentralization of required resources and contents near users and at the edge. A crucial challenge in this direction is the development of a smart mechanism to effectively cache contents upon Edge Servers (ESs) near users for high effectiveness and low latency of content delivery with the constraints on computational and storage capacities of ESs. This study employs a queuing model for analyzing total request delay and interprets the content caching problem as an adversarial semi-bandits problem. We propose an Online Self-feedback Adversarial Semi-bandits Learning (OSAL) algorithm that incorporates a dual-layer learning architecture for dynamically generating caching strategies and maximizes the long-term reward. Experimental results demonstrate that the proposed method significantly outperforms the state-of-the-art methods across various performance metrics in a real-world multi mobile-user content caching case.


  • Open Access

    Personalized Explainable Recommendations for Self-Attention Collaboration

    by Yongfu Zha, Xuanxuan Che, Lina Sun, Yumin Dong

    In recommender systems, providing reasonable explanations can enhance users’ comprehension of recommended results. Template-based explainable recommendation heavily relies on pre-defined templates, constraining the expressiveness of generated sentences and resulting in low-quality explanations. Recently, a novel approach was introduced, utilizing embedding representations of items and comments to address the issue of user IDs and item IDs not residing in the same semantic space as words, thus attributing linguistic meaning to IDs. However, these models often fail to fully exploit collaborative information within the data. In personalized recommendation and explanation processes, understanding the user’s emotional feedback and feature preferences is paramount. To address this, we propose a personalized explainable recommendation model based on self-attention collaboration. Initially, the model employs an attention network to amalgamate the user’s historical interaction feature preferences with their user ID information, while simultaneously integrating all feature information of the item with its item ID to enhance semantic ID representation. Subsequently, the model incorporates the user’s comment feature rhetoric and sentiment feedback to generate more personalized recommendation explanations utilizing a self-attention network. Experimental evaluations conducted on two datasets of varying scales demonstrate the superiority of our model over current state-of-the-art approaches, validating its effectiveness.

  • Open Access

    Research of the Variational Shadow Quantum Circuit Based on the Whale Optimization Algorithm in Image Classification

    by Shuang Wu, Xueliang Song, Yumin Dong, Fanghua Jia

    In order to explore the possibility of cross-fertilization between quantum computing and neural networks as well as to improve the classification performance of quantum neural networks, this paper proposes an improved Variable Split Shadow Quantum Circuit (VSQC-WOA) model based on the Whale Optimization Algorithm. In this model, we design a strongly entangled local shadow circuit to achieve efficient characterization of global features through local shadow feature extraction and a sliding mechanism, which provides a rich quantum feature representation for the classification task. The gradient is then computed by the parameter-shifting method, and finally the features processed by the shadow circuit are passed to the classical fully connected neural network (FCNN) for processing and classification. The model also introduces the Whale Optimization Algorithm (WOA) to further optimize the weights and biases of the fully connected neural network, which improves the expressive power and classification accuracy of the model. In this paper, we firstly use different localized shadow circuit VSQC models to achieve the binary classification task on the MNIST dataset, and our design of strongly entangled shadow circuits performs the best in terms of classification accuracy. The VSQC-WOA model is then used to multi-classify the MNIST dataset (three classifications as an example), and the effectiveness of the proposed VSQC-WOA model as well as the robustness and generalization ability of the model are verified through various comparison experiments.

  • Open Access

    A quantum image encryption scheme based on quantum Arnold transform and hyper 5D chaotic system

    by Lishi Liu, Chenhao Yin, Yumin Dong

    In the wake of the information security era, ensuring the privacy, accuracy, and reliability of images has become imperative. In this paper, a quantum image encryption method combining quantum Arnold transform and a hyper 5D chaotic system is devised. Furthermore, first, a nonlinear color quantization model (NCQI) is used to convert digital images in standard RGB format into quantum color images. Then, based on the size of the converted quantum color image, the corresponding encryption key is generated using the hyper 5D chaotic system. In the encryption stage, we use the key to encrypt the converted quantum color image. This process entails two principal steps: First, we use quantum Arnold transform to scramble the pixel positions in the image and, subsequently, diffuse the pixel values through a carefully designed cyclic shift strategy to achieve the preliminary encryption of the image. To further strengthen the encryption security, we use a quantum rotation gate to perform quantum rotation transformation on the color channel of each pixel point in the encrypted image separately. These well-designed encryption processes, in succession, significantly enhance the security of the image. This paper provides the corresponding quantum circuit design and validates both the efficacy and security of the quantum image encryption scheme through simulation experimental results.


  • Research on medical image denoising using an improved dung beetle optimization algorithm based on a novel 3D chaotic system

    by Yumin Dong, Tonglei Sun

    Artifacts and noise not only reduce the imaging quality of medical images but also threaten the integrity and accuracy of medical data features. Therefore, image denoising has become an indispensable step in the medical imaging process. However, previous medical image denoising methods have shown certain limitations when applied to different types of images. For example, using filtering methods for image denoising often results in edge blurring and even loss of image details. This significantly affects the integrity and accuracy of medical data features when processing detail-rich medical images, such as brain images.To address these limitations, this paper uses the improved dung beetle optimization algorithm of 3D chaotic system combined with hybrid filter to achieve better denoising effect. The newly proposed 3D chaotic system (3D-GP) enhances the randomness of the initial population position in the dung beetle optimization algorithm, ensuring that the search space follows a symmetric probabilistic distribution, thus preventing the DBO algorithm from falling into local optima. The improved algorithm exhibits a high degree of adaptability, dynamically adjusting parameter settings based on the characteristics of different medical images to achieve more consistent and stable denoising results. Based on the improved DBO algorithm, the incorporation of hybrid filters enables effective handling of various types of noise. Specifically, for complex medical images with speckle noise and Gaussian noise, the proposed algorithm demonstrates remarkable denoising effectiveness. Experimental results show that the improved DBO algorithm can effectively remove noise while preserving image details, especially when processing images with subtle structures and strong noise interference. Compared with traditional filtering methods, the proposed algorithm not only improves the imaging quality of medical images but also ensures the integrity and accuracy of medical data features.