Searcharxiv⌕ Search

arXiv subjects

Zheng Lu

Publications and source records attributed to Zheng Lu.

At least 37 records · Page 2Linked to original sources
arXiv2026

DIRCR: Dual-Inference Rule-Contrastive Reasoning for Solving RAVENs

Abstract visual reasoning remains challenging as existing methods often prioritize either global context or local row-wise relations, failing to integrate both, and lack intermediate feature constraints, leading to incomplete rule capture and entangled representations. To address these issues, we propose the Dual-Inference Rule-Contrastive Reasoning (DIRCR) model. Its core component, the Dual-Inference Reasoning Module, combines a local path for row-wise analogical reasoning and a global path for holistic inference, integrated via a gated attention mechanism. Additionally, a Rule-Contrastive Learning Module introduces pseudo-labels to construct positive and negative rule samples, applying contrastive learning to enhance feature separability and promote abstract, transferable rule learning. Experimental results on three RAVEN datasets demonstrate that DIRCR significantly enhances reasoning robustness and generalization. Codes are available at https://github.com/csZack-Zhang/DIRCR.

cs.AI↗
arXiv2026

Vision KAN: Towards an Attention-Free Backbone for Vision with Kolmogorov-Arnold Networks

Attention mechanisms have become a key module in modern vision backbones due to their ability to model long-range dependencies. However, their quadratic complexity in sequence length and the difficulty of interpreting attention weights limit both scalability and clarity. Recent attention-free architectures demonstrate that strong performance can be achieved without pairwise attention, motivating the search for alternatives. In this work, we introduce Vision KAN (ViK), an attention-free backbone inspired by the Kolmogorov-Arnold Networks. At its core lies MultiPatch-RBFKAN, a unified token mixer that combines (a) patch-wise nonlinear transform with Radial Basis Function-based KANs, (b) axis-wise separable mixing for efficient local propagation, and (c) low-rank global mapping for long-range interaction. Employing as a drop-in replacement for attention modules, this formulation tackles the prohibitive cost of full KANs on high-resolution features by adopting a patch-wise grouping strategy with lightweight operators to restore cross-patch dependencies. Experiments on ImageNet-1K show that ViK achieves competitive accuracy with linear complexity, demonstrating the potential of KAN-based token mixing as an efficient and theoretically grounded alternative to attention.

cs.CV↗
arXiv2026

R-PINN: Recovery-type a-posteriori estimator enhanced adaptive PINN

In recent years, with the advancements in machine learning and neural networks, algorithms using physics-informed neural networks (PINNs) to solve PDEs have gained widespread applications. While these algorithms are well-suited for a wide range of equations, they often exhibit suboptimal performance when applied to equations with large local gradients, resulting in substantial localized errors. To address this issue, this paper proposes an adaptive PINN algorithm designed to improve accuracy in such cases. The core idea of the algorithm is to adaptively adjust the distribution of collocation points based on the recovery-type a-posterior error of the current numerical solution, enabling a better approximation of the true solution. This approach is inspired by the adaptive finite element method. By combining the recovery-type a-posteriori estimator, a gradient-recovery estimator commonly used in the adaptive finite element method (FEM) with PINNs, we introduce the Recovery-type a-posteriori estimator enhanced adaptive PINN (R-PINN) and compare its performance with a typical adaptive PINN algorithm, FI-PINN. Our results demonstrate that R-PINN achieves faster convergence with fewer adaptive points and significantly outperforms in the cases with multiple regions of large errors than FI-PINN. Notably, our method is a hybrid numerical approach for solving partial differential equations, integrating adaptive FEM with PINNs.

math.NA↗
arXiv2025

MMCR: Benchmarking Cross-Source Reasoning in Scientific Papers

Fully comprehending scientific papers by machines reflects a high level of Artificial General Intelligence, requiring the ability to reason across fragmented and heterogeneous sources of information, presenting a complex and practically significant challenge. While Vision-Language Models (VLMs) have made remarkable strides in various tasks, particularly those involving reasoning with evidence source from single image or text page, their ability to use cross-source information for reasoning remains an open problem. This work presents MMCR, a high-difficulty benchmark designed to evaluate VLMs' capacity for reasoning with cross-source information from scientific papers. The benchmark comprises 276 high-quality questions, meticulously annotated by humans across 7 subjects and 10 task types. Experiments with 18 VLMs demonstrate that cross-source reasoning presents a substantial challenge for existing models. Notably, even the top-performing model, GPT-4o, achieved only 48.55% overall accuracy, with only 20% accuracy in multi-table comprehension tasks, while the second-best model, Qwen2.5-VL-72B, reached 39.86% overall accuracy. Furthermore, we investigated the impact of the Chain-of-Thought (CoT) technique on cross-source reasoning and observed a detrimental effect on small models, whereas larger models demonstrated substantially enhanced performance. These results highlight the pressing need to develop VLMs capable of effectively utilizing cross-source information for reasoning.

cs.CL↗
arXiv2025

MedKAN: An Advanced Kolmogorov-Arnold Network for Medical Image Classification

Recent advancements in deep learning for image classification predominantly rely on convolutional neural networks (CNNs) or Transformer-based architectures. However, these models face notable challenges in medical imaging, particularly in capturing intricate texture details and contextual features. Kolmogorov-Arnold Networks (KANs) represent a novel class of architectures that enhance nonlinear transformation modeling, offering improved representation of complex features. In this work, we present MedKAN, a medical image classification framework built upon KAN and its convolutional extensions. MedKAN features two core modules: the Local Information KAN (LIK) module for fine-grained feature extraction and the Global Information KAN (GIK) module for global context integration. By combining these modules, MedKAN achieves robust feature modeling and fusion. To address diverse computational needs, we introduce three scalable variants--MedKAN-S, MedKAN-B, and MedKAN-L. Experimental results on nine public medical imaging datasets demonstrate that MedKAN achieves superior performance compared to CNN- and Transformer-based models, highlighting its effectiveness and generalizability in medical image analysis.

cs.CV↗
arXiv2024

Activation Space Selectable Kolmogorov-Arnold Networks

The multilayer perceptron (MLP), a fundamental paradigm in current artificial intelligence, is widely applied in fields such as computer vision and natural language processing. However, the recently proposed Kolmogorov-Arnold Network (KAN), based on nonlinear additive connections, has been proven to achieve performance comparable to MLPs with significantly fewer parameters. Despite this potential, the use of a single activation function space results in reduced performance of KAN and related works across different tasks. To address this issue, we propose an activation space Selectable KAN (S-KAN). S-KAN employs an adaptive strategy to choose the possible activation mode for data at each feedforward KAN node. Our approach outperforms baseline methods in seven representative function fitting tasks and significantly surpasses MLP methods with the same level of parameters. Furthermore, we extend the structure of S-KAN and propose an activation space selectable Convolutional KAN (S-ConvKAN), which achieves leading results on four general image classification datasets. Our method mitigates the performance variability of the original KAN across different tasks and demonstrates through extensive experiments that feedforward KANs with selectable activations can achieve or even exceed the performance of MLP-based methods. This work contributes to the understanding of the data-centric design of new AI paradigms and provides a foundational reference for innovations in KAN-based network architectures.

cs.LG↗
arXiv2024

Helicity-conservative Physics-informed Neural Network Model for Navier-Stokes Equations

We design the helicity-conservative physics-informed neural network model for the Navier-Stokes equation in the ideal case. The key is to provide an appropriate PDE model as loss function so that its neural network solutions produce helicity conservation. Physics-informed neural network model is based on the strong form of PDE. We compare the proposed Physics-informed neural network model and a relevant helicity-conservative finite element method. We arrive at the conclusion that the strong form PDE is better suited for conservation issues. We also present theoretical justifications for helicity conservation as well as supporting numerical calculations.

physics.comp-ph↗
arXiv2023

Rigidity and quantitative stability for partially overdetermined problems and capillary CMC hypersurfaces

In this paper, we first prove a rigidity result for a Serrin-type partially overdetermined problem in the half-space, which gives a characterization of capillary spherical caps by the overdetermined problem. In the second part, we prove quantitative stability results for the Serrin-type partially overdetermined problem, as well as capillary almost constant mean curvature hypersurfaces in the half-space.

math.AP↗
arXiv2023

Capillary Schwarz symmetrization in the half-space

In this paper, we introduce a notion of capillary Schwarz symmetrization in the half-space. It can be viewed as the counterpart of the classical Schwarz symmetrization in the framework of capillary problem in the half-space. A key ingredient is a special anisotropic gauge, which enables us to transform the capillary symmetrization to the convex symmetrization introduced in \cite{AFT97}.

math.AP↗
arXiv2022

The future of blockchain-enabled tokenization in infrastructure investment and development: A Delphi-based scenario analysis

Spurred by the emerging blockchain technology and increased interest in tokenization, this forecasting research built on extensive literature and aggregated expertise to explore the potential implementation of blockchain-enabled tokenization in infrastructure investment and development. The Delphi-based scenario analysis approach was applied to gather long-term forecasts and assessments of a research panel consisting of 39 experts in blockchain tokenization and infrastructure development on how tokenization will influence the future of infrastructure finance and identify scenarios of potential applications and impact. International experts were segregated into two groups salient to this topical area based on both experience ad self-identification: infrastructure development and blockchain tokenization. Twenty-three projections for 2035, developed from a literature review, case study analysis, and expert interviews, concerning perspectives of both the supply and demand side for the adoption of blockchain tokenization, were assessed in a two-round Delphi analysis. Regulatory, economic, social, and technological perspectives of tokenization were taken into consideration. Assessments were based on both probability and impact of occurrence. Three groups of scenarios resulted from quantitative and qualitative analysis, reflecting agreement and differentiation between both expert groups. The results of this study clearly underlined the potential of tokenization in infrastructure. Uncertainties and barriers confronting the technologies' diffusion were discussed. This study contributes to the transfer of general technical-driven blockchain-enabled tokenization knowledge to infrastructure-specific tokenization knowledge. Long-term strategic planning is supported by this study with the scenario data acting as a starting point for blockchain-related efforts in infrastructure development.

cs.CY↗
arXiv2022

Low Latency Techniques for Mobile Backhaul over DOCSIS

The mobile network operators (MNOs) are looking into economically viable backhaul solutions as alternatives to fiber, specifically the hybrid fiber coaxial networks (HFC). When the latencies from both the wireless and the HFC networks are added together, the result is a noticeable end-to-end system latency, particularly under network congestion. In order to decrease total system latency, we proposed a method to improve upstream user- to-mobile core latency by coordinating the LTE and HFC scheduling in previous papers. In this paper, we implement and optimize the proposed method on a custom LTE and DOCSIS end-to-end system testbed. The testbed uses the OpenAirInterface (OAI) platform for the LTE network, along with Cisco's broadband router cBR-8 that is currently deployed in the HFC networks around the world. Our results show a backhaul latency improvement under all traffic load conditions.

cs.NI↗
arXiv2022

Improving Visual-Semantic Embedding with Adaptive Pooling and Optimization Objective

Visual-Semantic Embedding (VSE) aims to learn an embedding space where related visual and semantic instances are close to each other. Recent VSE models tend to design complex structures to pool visual and semantic features into fixed-length vectors and use hard triplet loss for optimization. However, we find that: (1) combining simple pooling methods is no worse than these sophisticated methods; and (2) only considering the most difficult-to-distinguish negative sample leads to slow convergence and poor Recall@K improvement. To this end, we propose an adaptive pooling strategy that allows the model to learn how to aggregate features through a combination of simple pooling methods. We also introduce a strategy to dynamically select a group of negative samples to make the optimization converge faster and perform better. Experimental results on Flickr30K and MS-COCO demonstrate that a standard VSE using our pooling and optimization strategies outperforms current state-of-the-art systems (at least 1.0% on the metrics of recall) in image-to-text and text-to-image retrieval. Source code of our experiments is available at https://github.com/96-Zachary/vse_2ad.

cs.MM↗
arXiv2021

A Multilayer Network Model of the Coevolution of the Spread of a Disease and Competing Opinions

During the COVID-19 pandemic, conflicting opinions on physical distancing swept across social media, affecting both human behavior and the spread of COVID-19. Inspired by such phenomena, we construct a two-layer multiplex network for the coupled spread of a disease and conflicting opinions. We model each process as a contagion. On one layer, we consider the concurrent evolution of two opinions -- pro-physical-distancing and anti-physical-distancing -- that compete with each other and have mutual immunity to each other. The disease evolves on the other layer, and individuals are less likely (respectively, more likely) to become infected when they adopt the pro-physical-distancing (respectively, anti-physical-distancing) opinion. We develop approximations of mean-field type by generalizing monolayer pair approximations to multilayer networks; these approximations agree well with Monte Carlo simulations for a broad range of parameters and several network structures. Through numerical simulations, we illustrate the influence of opinion dynamics on the spread of the disease from complex interactions both between the two conflicting opinions and between the opinions and the disease. We find that lengthening the duration that individuals hold an opinion may help suppress disease transmission, and we demonstrate that increasing the cross-layer correlations or intra-layer correlations of node degrees may lead to fewer individuals becoming infected with the disease.

cs.SI↗
arXiv2021

Analytics and Machine Learning in Vehicle Routing Research

The Vehicle Routing Problem (VRP) is one of the most intensively studied combinatorial optimisation problems for which numerous models and algorithms have been proposed. To tackle the complexities, uncertainties and dynamics involved in real-world VRP applications, Machine Learning (ML) methods have been used in combination with analytical approaches to enhance problem formulations and algorithmic performance across different problem solving scenarios. However, the relevant papers are scattered in several traditional research fields with very different, sometimes confusing, terminologies. This paper presents a first, comprehensive review of hybrid methods that combine analytical techniques with ML tools in addressing VRP problems. Specifically, we review the emerging research streams on ML-assisted VRP modelling and ML-assisted VRP optimisation. We conclude that ML can be beneficial in enhancing VRP modelling, and improving the performance of algorithms for both online and offline VRP optimisations. Finally, challenges and future opportunities of VRP research are discussed.

cs.LG↗
arXiv2020

Learning Regular Expressions for Interpretable Medical Text Classification Using a Pool-based Simulated Annealing and Word-vector Models

In this paper, we propose a rule-based engine composed of high quality and interpretable regular expressions for medical text classification. The regular expressions are auto generated by a constructive heuristic method and optimized using a Pool-based Simulated Annealing (PSA) approach. Although existing Deep Neural Network (DNN) methods present high quality performance in most Natural Language Processing (NLP) applications, the solutions are regarded as uninterpretable black boxes to humans. Therefore, rule-based methods are often introduced when interpretable solutions are needed, especially in the medical field. However, the construction of regular expressions can be extremely labor-intensive for large data sets. This research aims to reduce the manual efforts while maintaining high-quality solutions

cs.CL↗
arXiv2020

Retrieving and ranking short medical questions with two stages neural matching model

Internet hospital is a rising business thanks to recent advances in mobile web technology and high demand of health care services. Online medical services become increasingly popular and active. According to US data in 2018, 80 percent of internet users have asked health-related questions online. Numerous data is generated in unprecedented speed and scale. Those representative questions and answers in medical fields are valuable raw data sources for medical data mining. Automated machine interpretation on those sheer amount of data gives an opportunity to assist doctors to answer frequently asked medical-related questions from the perspective of information retrieval and machine learning approaches. In this work, we propose a novel two-stage framework for the semantic matching of query-level medical questions.

cs.CL↗
arXiv2020

Micromobility in Smart Cities: A Closer Look at Shared Dockless E-Scooters via Big Social Data

The micromobility is shaping first- and last-mile travels in urban areas. Recently, shared dockless electric scooters (e-scooters) have emerged as a daily alternative to driving for short-distance commuters in large cities due to the affordability, easy accessibility via an app, and zero emissions. Meanwhile, e-scooters come with challenges in city management, such as traffic rules, public safety, parking regulations, and liability issues. In this paper, we collected and investigated 5.8 million scooter-tagged tweets and 144,197 images, generated by 2.7 million users from October 2018 to March 2020, to take a closer look at shared e-scooters via crowdsourcing data analytics. We profiled e-scooter usages from spatial-temporal perspectives, explored different business roles (i.e., riders, gig workers, and ridesharing companies), examined operation patterns (e.g., injury types, and parking behaviors), and conducted sentiment analysis. To our best knowledge, this paper is the first large-scale systematic study on shared e-scooters using big social data.

cs.SI↗