The Dung beetle optimization (DBO) algorithm, devised by Jiankai Xue in 2022, is known for its strong optimization capabilities and fast convergence. However, it does have certain limitations, including insufficiently random population initialization, slow search speed, and inadequate global search capabilities. Drawing inspiration from the mathematical properties of the Sinh and Cosh functions, we proposed a new metaheuristic algorithm, Sinh–Cosh Dung Beetle Optimization (SCDBO). By leveraging the Sinh and Cosh functions to disrupt the initial distribution of DBO and balance the development of rollerball dung beetles, SCDBO enhances the search efficiency and global exploration capabilities of DBO through nonlinear enhancements. These improvements collectively enhance the performance of the dung beetle optimization algorithm, making it more adept at solving complex real-world problems. To evaluate the performance of the SCDBO algorithm, we compared it with seven typical algorithms using the CEC2017 test functions. Additionally, by successfully applying it to three engineering problems, robot arm design, pressure vessel problem, and unmanned aerial vehicle (UAV) path planning, we further demonstrate the superiority of the SCDBO algorithm.

Phylogenetic analyses of genome sequences from infectious pathogens reveal essential information regarding their evolution and transmission, as seen during the coronavirus disease 2019 pandemic. Recently developed pandemic-scale phylogenetic inference methods reduce the computational demand of phylogenetic reconstruction from genomic epidemiological datasets, allowing the analysis of millions of closely related genomes. However, widespread homoplasies, due to recurrent mutations and sequence errors, cause phylogenetic uncertainty and biases. We present algorithms and models to substantially improve the computational performance and accuracy of pandemic-scale phylogenetics. In particular, we account for, and identify, mutation rate variation and recurrent sequence errors. We reconstruct a reliable and public sequence alignment and phylogenetic tree of >2 million severe acute respiratory syndrome coronavirus 2 genomes encapsulating the evolutionary history and global spread of the virus up to February 2023.

To enhance the convergence efficiency and solution precision of the Red-billed Blue Magpie Optimizer (RBMO), this study proposes a Multi-Strategy Enhanced Red-billed Blue Magpie Optimizer (MRBMO). The principal methodological innovations encompass three aspects: (1) Development of a novel dynamic boundary constraint handling mechanism that strengthens algorithmic exploration capabilities through adaptive regression strategy adjustment for boundary-transgressing particles; (2) Incorporation of an elite guidance strategy during the predation phase, establishing a guided search framework that integrates historical individual optimal information while employing a Lévy Flight strategy to modulate search step sizes, thereby achieving effective balance between global exploration and local exploitation capabilities; (3) Comprehensive experimental evaluations conducted on the CEC2017 and CEC2022 benchmark test suites demonstrate that MRBMO significantly outperforms classical enhanced algorithms and exhibits competitive performance against state-of-the-art optimizers across 41 standardized test functions. The practical efficacy of the algorithm is further validated through successful applications to four classical engineering design problems, confirming its robust problem-solving capabilities.

Convolutional neural networks (CNNs) are a representative class of deep learning algorithms including convolutional computation that perform translation invariant classification of input data based on their hierarchical architecture. However, classical convolutional neural network learning methods use the steepest descent algorithm for training, and the learning performance is greatly influenced by the initial weight settings of the convolutional and fully connected layers, requiring re-tuning to achieve better performance under different model structures and data. Combining the strengths of the simulated annealing algorithm in global search, we propose applying it to the hyperparameter search process in order to increase the effectiveness of convolutional neural networks (CNNs). In this paper, we introduce SA-CNN neural networks for text classification tasks based on Text-CNN neural networks and implement the simulated annealing algorithm for hyperparameter search. Experiments demonstrate that we can achieve greater classification accuracy than earlier models with manual tuning, and the improvement in time and space for exploration relative to human tuning is substantial.

In broad terms, accessibility measures opportunities (e.g., shops, residents) reachable within a given time frame. Urban rail transit (URT) plays a crucial role in providing accessibility, but it is susceptible to disruptions. In city centers with dense public transport (PT) networks, travelers can often find alternative lines. However, in suburbs where PT is sparse, disruptions have a more significant impact on accessibility. The traditional approach consists in deploying bridge and replacement buses to mitigate URT disruptions without specific care to accessibility. Yet, the question arises: is this approach the most effective way to restore accessibility? To the best of our knowledge, our paper is the first to propose a bus re-routing method with the objective of restoring accessibility during URT disruptions. We formulate an integer program and develop a two-stage heuristic algorithm to maximize restored accessibility. The efficacy of our method is always the present assessed in Évry-Courcouronnes and Choisy-le-Roi, France. The results show that, compared with conventional replacement methods, our strategy improves accessibility, in particular in the areas most affected by the disruption. Such results are observed even when no additional vehicles are deployed and, at the same time, achieve a reduction in the kilometers traveled. Despite it being well understood that accessibility is the most relevant benefit a transportation system can produce, this aspect is reflected by the traditional approaches in remediation to disruption. With this work, we show instead how to make accessibility the main guiding principle in remediation.