News Center
News Center

Young Faculty Member Xue Guangdong from the School of Mathematics and Statistics Makes Progress in Computational Intelligence Research

2025-02-28

Recently, Dr. Xue Guangdong, a young faculty member from the School of Mathematics and Statistics at Donghua University, has made significant research progress in the field of computational intelligence.

Dr. Xue published an academic paper titled “ADMTSK: A High-Dimensional Takagi-Sugeno-Kang Fuzzy System Based on Adaptive Dombi T-Norm” in the internationally renowned journal IEEE Transactions on Fuzzy Systems (IEEE TFS). Dr. Xue is the first author of the paper, and Donghua University is the first affiliation. This research was supported by the Fundamental Research Funds for the Central Universities.

High-dimensional data (datasets with thousands or even tens of thousands of features) are becoming increasingly common across various fields, posing an inevitable challenge in industrial applications: how to effectively process such data. Fuzzy systems, a widely used intelligent method in computational intelligence, face significant challenges in handling high-dimensional data due to inherent limitations. To address this meaningful problem, the paper explores techniques for effectively applying fuzzy systems to high-dimensional data by designing and improving key operators such as membership functions and T-norms, ultimately proposing a novel high-dimensional fuzzy system modeling approach.

First, the author developed a fuzzy system modeling method based on the Dombi T-norm, tailored for high-dimensional data. This approach is both simple and effective. To further enhance its performance, an adaptive strategy was introduced for the indicative parameters of the Dombi T-norm, for which a composite Gaussian membership function was also proposed. This resulted in a high-dimensional fuzzy system modeling approach based on the adaptive Dombi T-norm. Numerical experiments demonstrated that this method achieved satisfactory recognition accuracy even for problems involving tens of thousands of features, outperforming similar methods. This research alleviates the limitations of fuzzy systems in processing high-dimensional data, expands the application scenarios of fuzzy intelligent models, and makes a positive impact on the development of the fuzzy systems field.