Anhui Polytechnic University (AHPU, Chinese: 安徽工程大学; pinyin: Ānhuī gōngchéng daxue) is a comprehensive public university based in Wuhu prefecture level city, Anhui province, China. The university provides programmes in the fields of science and technology, engineering, management and economics.[non-primary source needed].
The question of which graphs are determined by their spectra is a central problem in spectral graph theory. In this paper, we investigate spectral determination for complement-reducible graphs, or cographs. By combining the Johnson-Newman theorem on generalized cospectrality with standard tools in the asymptotic enumeration of trees, we prove a Schwenk-type result for cographs: almost all cographs have a generalized cospectral mate within the family of cographs, and hence have a cospectral mate. This contrasts sharply with threshold graphs, a well-studied subclass of cographs, where no threshold graph has a cospectral mate within that family. We also establish a signless Laplacian analogue by proving that almost all cographs have a generalized Q-cospectral mate. (c) 2026 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
Continual multi-view clustering (CMVC) aims to discover cluster structures from multi-view data that arrive sequentially and exhibit distribution shifts. However, existing incremental anchor-based frameworks for CMVC still have some problems: 1. Anchors are learned in a purely reconstruction-driven way, resulting in ambiguous cluster semantics and higher sensitivity to noise. 2. Existing methods overlook the role of consistent anchor-level semantics, which limits their ability to suppress semantic drift. 3. It is tough to align the anchor graphs, weakening temporal consistency and degrading clustering stability. To address these limitations, we propose a novel continual multi-view clustering method with pseudo-label guided consistency anchor learning (CMVC-PCAL). In the initialization stage, we build a unified anchor graph representation by decomposing each anchor graph into a shared consensus component and a view-specific inconsistency component, and by introducing cluster-wise pseudo labels to align anchors with latent clusters, yielding a clearer and more noise-robust consensus anchor graph. In the continual stage, for each new data batch, a lightweight incremental scheme updates anchors and anchor graphs: an alignment matrix aligns historical and current anchors, and a consistency regularization term aligns their consensus parts, while a pseudo-label guided term jointly updates anchor pseudo labels and cluster centers to continuously refine anchor-level semantics. Finally, an alternating optimization algorithm is developed to efficiently train the model, and experiments on public multi-view benchmark datasets verify its effectiveness and efficiency in continual multi-view clustering scenarios.
In soft sensing of multi-mode chemical processes with scarce labeled data, existing methods often overlook the spatiotemporal coupling characteristics of process data, and models tend to suffer from performance degradation and cross-mode adaptation capability limitation under continuous mode switching scenarios. This article proposes a continual transfer graph learning (CTGL) model for soft sensing in multimode processes. First, a spatiotemporal graph model is constructed using historical data to capture the spatiotemporal coupling relationships among variables. When a new mode emerges, a Bayesian-based continual transfer framework is established, elastic weight consolidation retains important information from historical modes, while an active forgetting mechanism discards useless information from historical modes that is detrimental to modeling under the new mode. Together, both mechanisms constrain the fine-tuning process for the new mode. Moreover, a multi-output head strategy is incorporated to retain prediction tasks of historical modes while enabling rapid adaptation to the prediction task of the new mode, thereby enhancing the transfer capability of the CTGL model under multimode switching scenarios. The proposed method is validated in the penicillin fermentation and the wastewater treatment cases. Compared with compared baselines, CTGL reduces average root mean square error by 38.7% and improves average coefficient of determination by 5.1% in the penicillin fermentation case.
The proliferation of in-vehicle infotainment systems raises critical safety concerns regarding driver distraction. This driving simulator study provides empirical evidence on how interaction modalities (touchscreen, voice, gesture) for non-driving-related tasks (NDRTs) affect driver state under varying traffic density and lighting conditions. Results showed that touchscreen interaction was associated with higher workload-related EEG activity, reduced HRV, larger eyelid aperture, lower situation awareness (SA), and less stable vehicle-control measures. Nighttime driving was associated with larger pupil diameter, while some adverse modality-related outcomes were more pronounced under high-density nighttime conditions. In contrast, voice interaction and WoZ-implemented gesture interaction were associated with lower workload-related indicators under the present controlled simulator conditions. A layered SA analysis revealed that perception-level SA was primarily impaired by visually demanding touchscreen use, while comprehension- and projection-level SA were more affected by environmental complexity. Within the scope of this driving simulator study—focusing on a music-selection task among young drivers—the proposed multimodal assessment framework contributes empirical evidence that may inform situation-aware adaptive interface design and scenario-based evaluation approaches for integrated HMI systems. These findings caution against overreliance on touchscreen interfaces for music-selection tasks in complex driving environments and suggest the potential safety relevance of lower-visual-demand modalities such as voice interaction and controlled gesture interaction, while recognizing that production-grade gesture recognition may introduce additional recognition-related costs.
Urban co-pollution by PM₂.₅ and O₃ poses divergent public health and economic challenges. Using concentration-response functions and the value of statistical life (VSL) model, this study quantifies health and economic burdens across 41 YRD cities from 2014 to 2023 and identifies distinct drivers via two-way fixed effects regression. Key findings show that PM₂.₅-related premature deaths declined by 7.77