Author name cluster

Ning Yao

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

3 papers

2 author rows

EAAI Journal 2026 Journal Article

Fuzzy reasoning method based on intuitionistic fuzzy similarity measure and its application in pattern recognition

Ning Yao
Anni Zhang
Ruirui Zhao
Minxia Luo

The reducibility of a method and the similarity between the rule antecedent and the input variable are two extremely paramount factors in the construction of reasoning method. Therefore, this paper proposes a reducible reasoning method based on intuitionistic fuzzy similarity measure. In detail, the definition of intuitionistic fuzzy similarity measure being able to comprehensively portray information is first given via the intuitionistic residuated implication operator. With this definition, we discuss the intuitionistic fuzzy similarity reasoning method to solve the intuitionistic fuzzy modus ponens problem. Meanwhile, a series of properties for the proposed reasoning method are investigated. By introducing the intuitionistic fuzzy similarity reasoning method into the pattern recognition process, intuitionistic fuzzy similarity reasoning-based scheme for pattern recognition is developed, which can effectively solve the specific pattern recognition problems. Several experiments demonstrate the effectiveness and applicability of the presented reasoning method.

Details DOI

ICLR Conference 2025 Conference Paper

Learning Graph Quantized Tokenizers

Limei Wang
Kaveh Hassani
Si Zhang
Dongqi Fu
Baichuan Yuan
Weilin Cong
Zhigang Hua
Hao Wu

Transformers serve as the backbone architectures of Foundational Models, where domain-specific tokenizers allow them to adapt to various domains. Graph Transformers (GTs) have recently emerged as leading models in geometric deep learning, outperforming Graph Neural Networks (GNNs) in various graph learning tasks. However, the development of tokenizers for graphs has lagged behind other modalities, with existing approaches relying on heuristics or GNNs co-trained with Transformers. To address this, we introduce GQT (\textbf{G}raph \textbf{Q}uantized \textbf{T}okenizer), which decouples tokenizer training from Transformer training by leveraging multi-task graph self-supervised learning, yielding robust and generalizable graph tokens. Furthermore, the GQT utilizes Residual Vector Quantization (RVQ) to learn hierarchical discrete tokens, resulting in significantly reduced memory requirements and improved generalization capabilities. By combining the GQT with token modulation, a Transformer encoder achieves state-of-the-art performance on 20 out of 22 benchmarks, including large-scale homophilic and heterophilic datasets. The implementation is publicly available at \href{https://github.com/limei0307/GQT}{https://github.com/limei0307/GQT}.

Details

EAAI Journal 2024 Journal Article

Some novel Dice similarity measures for picture fuzzy sets and their applications

Ruirui Zhao
Zhangjie Zhou
Ning Yao
Minxia Luo

The similarity measures, as an important tool for assessing the similarity between picture fuzzy sets, have attracted the attention of numerous scholars and achieved rich achievements. Although existing similarity measures have exhibited their unique research contributions, they still fail to distinguish between highly similar but different picture fuzzy sets. This phenomenon poses a major challenge in applications, limiting our ability to make accurate judgments in scenarios that necessitate meticulous differentiation of picture fuzzy sets. Furthermore, the current research focuses on proposing individual and specific calculation formulas, particularly in the realm of Dice similarity measures. However, the systematic, uniform construction of Dice similarity measures remains mostly untapped. The main objective of this article is to propose a method for constructing Dice similarity measures. Through this approach, we can not only integrate existing Dice similarity measures into a unified framework, but also generate novel similarity measures. Firstly, we introduce a function that satisfies specific conditions as the basis for constructing Dice similarity measures. Secondly, by selecting different functions, we obtain various similarity measures. Finally, we propose the pattern recognition algorithm, the face recognition algorithm and the cluster analysis algorithm, all based on the proposed similarity measures. These algorithms are then applied to various fields, including pattern recognition, face recognition, and clustering analysis. Experimental results demonstrate that the proposed similarity measures not only compensate for the shortcomings of existing similarity measures but also exhibit high reliability and practicality, offering a powerful new tool for research and applications in related fields.

Details DOI