Author name cluster

Yifan Wei

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.

6 papers

2 author rows

EAAI Journal 2025 Journal Article

Adaptive learning network for detecting pavement distresses in complex environments

Dingfeng Wang
Allen A. Zhang
Yi Peng
Yifan Wei
Huixuan Cheng
Jing Shang

In pavement detection, simultaneously identifying multiple defects and surface design features can save time and reduce operational costs. ShuttleNet version 3 (slice-based), referred to as ShuttleNetS3, builds on ShuttleNet to enhance pixel-level pavement detection by addressing the challenge of identifying diverse defects and surface design features, particularly cracks, potholes, and other irregular issues. This improved architecture introduces key innovations, including a learnable down-sampling module, slice-based positional sampling, which uses slicing and positional encoding to capture diverse features, and a learnable up-sampling module, slice-based positional upsampling, which ensures accurate detail restoration through bi-directional mapping. Additionally, Deformable Convolution version 4 allows the model to adapt to morphological variations, enabling precise extraction of complex defect features. With an average F1 Score of 97. 48 % and an Intersection over Union of 95. 19 % on a test set of 1500 images, ShuttleNetS3 excels in detecting subtle and irregular defects. Its adaptive learning capability ensures optimal performance in dynamic environments, providing an intelligent and precise solution for road management and maintenance.

Details DOI

IJCAI Conference 2025 Conference Paper

SetKE: Knowledge Editing for Knowledge Elements Overlap

Yifan Wei
Xiaoyan Yu
Ran Song
Hao Peng
Angsheng Li

Large Language Models (LLMs) excel in tasks such as retrieval and question answering but require updates to incorporate new knowledge and reduce inaccuracies and hallucinations. Traditional updating methods, like fine-tuning and incremental learning, face challenges such as overfitting and high computational costs. Knowledge Editing (KE) provides a promising alternative but often overlooks the Knowledge Element Overlap (KEO) phenomenon, where multiple triplets share common elements, leading to editing conflicts. We identify the prevalence of KEO in existing KE datasets and show its significant impact on current KE methods, causing performance degradation in handling such triplets. To address this, we propose a new formulation, Knowledge Set Editing (KSE), and introduce SetKE, a method that edits sets of triplets simultaneously. Experimental results demonstrate that SetKE outperforms existing methods in KEO scenarios on mainstream LLMs. Additionally, we introduce EditSet, a dataset containing KEO triplets, providing a comprehensive benchmark.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in LLMs

Yifan Wei
Xiaoyan Yu
Tengfei Pan
Angsheng Li
Li Du

Large language models (LLMs) have achieved unprecedented performance by leveraging vast pretraining corpora, yet their performance remains suboptimal in knowledge-intensive domains such as medicine and scientific research, where high factual precision is required. While synthetic data provides a promising avenue for augmenting domain knowledge, existing methods frequently generate redundant samples that do not align with the model’s true knowledge gaps. To overcome this limitation, we propose a novel Structural Entropy-guided Knowledge Navigator (SENATOR) framework that addresses the intrinsic knowledge deficiencies of LLMs. Our approach employs the Structure Entropy (SE) metric to quantify uncertainty along knowledge graph paths and leverages Monte Carlo Tree Search (MCTS) to selectively explore regions where the model lacks domain-specific knowledge. Guided by these insights, the framework generates targeted synthetic data for supervised fine-tuning, enabling continuous self-improvement. Experimental results on LLaMA-3 and Qwen2 across multiple domain-specific benchmarks show that SENATOR effectively detects and repairs knowledge deficiencies, achieving notable performance improvements.

PDF Details

AAAI Conference 2025 Conference Paper

Towards Effective, Efficient and Unsupervised Social Event Detection in the Hyperbolic Space

Xiaoyan Yu
Yifan Wei
Shuaishuai Zhou
Zhiwei Yang
Li Sun
Hao Peng
Liehuang Zhu
Philip S. Yu

The vast, complex, and dynamic nature of social message data has posed challenges to social event detection (SED). Despite considerable effort, these challenges persist, often resulting in inadequately expressive message representations (ineffective) and prolonged learning durations (inefficient). In response to the challenges, this work introduces an unsupervised framework, HyperSED (Hyperbolic SED). Specifically, the proposed framework first models social messages into semantic-based message anchors, and then leverages the structure of the anchor graph and the expressiveness of the hyperbolic space to acquire structure- and geometry-aware anchor representations. Finally, HyperSED builds the partitioning tree of the anchor message graph by incorporating differentiable structural information as the reflection of the detected events. Extensive experiments on public datasets demonstrate HyperSED's competitive performance, along with a substantial improvement in efficiency compared to the current state-of-the-art unsupervised paradigm. Statistically, HyperSED boosts incremental SED by an average of 2%, 2%, and 25% in NMI, AMI, and ARI, respectively; enhancing efficiency by up to 37.41 times and at least 12.10 times, illustrating the advancement of the proposed framework.

PDF Details DOI

EAAI Journal 2023 Journal Article

Solving combined economic and emission dispatch problems using reinforcement learning-based adaptive differential evolution algorithm

Wenguan Luo
Xiaobing Yu
Yifan Wei

Nowadays, economic and environmental concerns in production have become increasingly significant. To address these issues, the Combined Economic and Emission Dispatch (CEED) problem has been introduced to optimize the power generation process by considering fuel cost and emitted substances. However, due to the nonlinearity and nonconvexity of the objective function, the optimization of CEED remains a challenge. In this paper, we develop a Reinforcement Learning-based Adaptive Differential Evolution (RLADE) algorithm to enhance the optimization performance. The mutation strategy and crossover probability of RLADE are optimized using Reinforcement Learning (RL) to respectively ensure better convergence speed and searchability. Additionally, two modifications of RL, namely the adaptive population size-based state division and fitness-ranking-based reward mechanism, are proposed to improve the accuracy of state division and reward calculation in RL. The experiments conducted in this paper consider two objective formulation methods of CEED problems, namely the quadratic and cubic criterion functions. The mean values and standard deviations of the obtained solutions were utilized to assess the performance of RLADE, as well as other comparative algorithms, namely DE algorithm and two RL-based DE variants. The results clearly demonstrate that RLADE surpasses its counterparts with proportion of 100%, 85. 7%, and 100% for the 6-unit and 11-unit quadratic CEED problems, as well as cubic criterion functions, in terms of both search accuracy and convergence ability. Furthermore, the significance of RLADE's superiority is confirmed through the Wilcoxon's signed rank test.

Details DOI

ECAI Conference 2020 Conference Paper

BTDE: Block Term Decomposition Embedding for Link Prediction in Knowledge Graph

Tao Luo 0010
Yifan Wei
Mei Yu 0004
Xuewei Li 0001
Mankun Zhao
Tianyi Xu
Jian Yu 0003
Jie Gao 0008

Link prediction is the main task of knowledge graph completion, predicting missing relations between entities based the existing links among the entities. The problem of knowledge graph completion can be framed as a third-order binary tensor completion problem. In this case, tensor decomposition seems like a natural solution. And many previous studies have shown that tensor decomposition methods are superior to Trans-based methods in link prediction experiments. Typical tensor decomposition methods are Canonical Polyadic (CP) decomposition and Tucker decomposition. In this paper, we propose Block term decomposition Embedding model (BTDE) for link prediction based on Block term decomposition (which can be seen as a combination of CP decomposition and Tucker decomposition) of the binary tensor representation of knowledge graph triples. The embeddings learned through BTDE is interpretable. In addition, we prove BTDE is fully expressive and derive the bound on its entity and relation embedding dimensionality for full expressivity which is the same as TuckER and smaller than the bound of previous start-of-the-art models ComplEx and SimplE. We show empirically that BTDE outperforms most previous state-of-the-art models across five standard link prediction datasets.

Details