The kernel method is a tool that converts data to a kernel space where operation can be performed. When converted to a high-dimensional feature space by using kernel functions, the data samples are more likely to be linearly separable. Traditional machine learning methods can be extended to the kernel space, such as the radial basis function (RBF) network. As a kernel-based method, support vector machine (SVM) is one of the most popular nonparametric classification methods, and is optimal in terms of computational learning theory. Based on statistical learning theory and the maximum margin principle, SVM attempts to determine an optimal hyperplane by addressing a quadratic programming (QP) problem. Using Vapnik–Chervonenkis dimension theory, SVM maximizes generalization performance by finding the widest classification margin within the feature space. In this paper, kernel machines and SVMs are systematically introduced. We first describe how to turn classical methods into kernel machines, and then give a literature review of existing kernel machines. We then introduce the SVM model, its principles, and various SVM training methods for classification, clustering, and regression. Related topics, including optimizing model architecture, are also discussed. We conclude by outlining future directions for kernel machines and SVMs. This article functions both as a state-of-the-art survey and a tutorial.

Ensemble learning and data fusion techniques play a crucial role in modern machine learning, enhancing predictive performance, robustness, and generalization. This paper provides a comprehensive survey of ensemble methods, covering foundational techniques such as bagging, boosting, and random forests, as well as advanced topics including multiclass classification, multiview learning, multiple kernel learning, and the Dempster–Shafer theory of evidence. We present a comparative analysis of ensemble learning and deep learning, highlighting their respective strengths, limitations, and synergies. Additionally, we examine the theoretical foundations of ensemble methods, including bias–variance trade-offs, margin theory, and optimization-based frameworks, while analyzing computational trade-offs related to training complexity, inference efficiency, and storage requirements. To enhance accessibility, we provide a structured comparative summary of key ensemble techniques. Furthermore, we discuss emerging research directions, such as adaptive ensemble methods, hybrid deep learning approaches, and multimodal data fusion, as well as challenges related to interpretability, model selection, and handling noisy data in high-stakes applications. By integrating theoretical insights with practical considerations, this survey serves as a valuable resource for researchers and practitioners seeking to understand the evolving landscape of ensemble learning and its future prospects.

Machine learning has become indispensable across various domains, yet understanding its theoretical underpinnings remains challenging for many practitioners and researchers. Despite the availability of numerous resources, there is a need for a cohesive tutorial that integrates foundational principles with state-of-the-art theories. This paper addresses the fundamental concepts and theories of machine learning, with an emphasis on neural networks, serving as both a foundational exploration and a tutorial. It begins by introducing essential concepts in machine learning, including various learning and inference methods, followed by criterion functions, robust learning, discussions on learning and generalization, model selection, bias–variance trade-off, and the role of neural networks as universal approximators. Subsequently, the paper delves into computational learning theory, with probably approximately correct (PAC) learning theory forming its cornerstone. Key concepts such as the VC-dimension, Rademacher complexity, and empirical risk minimization principle are introduced as tools for establishing generalization error bounds in trained models. The fundamental theorem of learning theory establishes the relationship between PAC learnability, Vapnik–Chervonenkis (VC)-dimension, and the empirical risk minimization principle. Additionally, the paper discusses the no-free-lunch theorem, another pivotal result in computational learning theory. By laying a rigorous theoretical foundation, this paper provides a comprehensive tutorial for understanding the principles underpinning machine learning.

Urban safety production is a core component of social safety and is associated with the production, storage and transport of hazardous chemicals, which are potential sources of disaster in an urban area. Chemicals’ locations in a city present a hidden site of danger, which can easily become disaster sites if supervision is inadequate. Aiming to improve the processes and typical scenarios of the production, storage, transportation and use of hazardous chemicals, this paper proposes an intelligent-technology-empowered active emergency command strategy (ITAECS) for urban hazardous chemical disaster management (UHCDM) in smart–safe cities. This paper aims to provide a strategy for active emergency command that takes into account the disaster source; hidden danger site; or disaster site of hazardous chemicals such as natural gas, gasoline and hydrogen energy based on five aspects: intelligent perception technology and equipment, a dynamically perceived IoT system, the accurate deduction of disaster posture, virtual reality emergency rescue rehearsal and an immersive emergency command platform. This research is conducive to the safety, efficiency and greenness of the whole industrial chain, such as the production, storage, transportation, operation and use of hazardous chemicals. There are difficulties and challenges in introducing ITAECS to urban hazardous chemical production safety and emergency management, such as the need for joint promotion of enterprises, industries and governments; uneven technological development; and several scientific–technological issues to be solved, as well as non-uniform standards. Overall, this paper helps improve the emergency management of urban hazardous chemical safety production.

This study investigates the influence of micro-arc oxidation (MAO) duration on the formation and characteristics of ceramic coatings on 6061 aluminum alloys. MAO treatments were conducted in a silicate-based electrolyte for varying durations. The effects of oxidation time on film thickness, surface roughness, morphology, and phase composition were meticulously assessed using a film thickness gauge, surface roughness meter, scanning electron microscopy (SEM), and X-ray diffraction (XRD) analysis. The corrosion behavior of the coatings was evaluated in a 3.5% NaCl solution using an electrochemical workstation. The experimental results demonstrate that both the film thickness and surface roughness exhibit a positive correlation with oxidation time. Notably, the film growth rate initially increased before plateauing, suggesting an optimal oxidation duration. The coatings primarily comprise γ-Al₂O₃, α-Al₂O₃, and Al phases, with the proportions of γ-Al₂O₃ and α-Al₂O₃ increasing progressively with extended oxidation times. Comparative electrochemical tests revealed that the treated specimens exhibited significantly higher corrosion potential and reduced corrosion current densities compared to the untreated aluminum alloy substrate. Specifically, a coating produced with a 20 min oxidation time demonstrated an exceptional corrosion current density (J_corr) of 1.545 × 10⁻⁶ A·cm⁻² and a corrosion resistance (R_p) of 2.716 × 10⁴ Ω·cm², signifying the best corrosion resistance achieved in this study. These results underscore the pivotal role of oxidation time in tailoring the properties of micro-arc-oxidized layers on 6061 aluminum alloys, offering valuable insights for enhancing corrosion protection in practical applications.

Aqueous zinc ion batteries (AZIBs) have emerged as a promising battery technology due to their excellent safety, high capacity, low cost, and eco-friendliness. However, the cycle life of AZIBs is limited by severe side reactions and zinc dendrite growth on the zinc electrode surface, hindering large-scale application. Here, an electrolyte optimization strategy utilizing the simplest dipeptide glycylglycine (Gly-Gly) additive is first proposed. Theoretical calculations and spectral analysis revealed that, due to the strong interaction between the amino group and Zn atoms, Gly-Gly preferentially adsorbs on zinc’s surface, constructing a stable and adaptive interfacial layer that inhibits zinc side reactions and dendrite growth. Furthermore, Gly-Gly can regulate zinc ion solvation, leading to a deposition mode shift from dendritic to lamellar and limiting two-dimensional dendrite diffusion. The symmetric cell with the addition of a 20 g/L Gly-Gly additive exhibits a cycle life of up to 1100 h. Under a high current density of 10 mA cm⁻², a cycle life of 750 cycles further demonstrates the reliable adaptability of the interfacial layer. This work highlights the potential of Gly-Gly as a promising solution for improving the performance of AZIBs.