In web spam detection, we propose a new semi-supervised learning algorithm named HFSSL (harmonic functions based learning). our method, labeled and unlabeled pages are represented as vertices in weighted graph. The problem is then modeled Gaussian random field on this graph, where the mean of characterized by harmonic functions, which can be efficiently obtained using matrix methods. experiments standard WEBSPAM-UK2006 show that effective.

This article is inspired by human's memory and recognition process to improve neural architecture search (NAS), which has shown novelty significance in the design of generative adversarial networks (GANs), but extremely enormous time consumption for searching GAN architectures based on reinforcement learning (RL) limits its applicability a great extent. The main reason behind challenge that, performance evaluation subnetworks during takes too much time. To solve this problem, we propose new...

Generative Adversarial Network (GAN), deemed as a powerful deep-learning-based silver bullet for intelligent data generation, has been widely used in multi-disciplines. Furthermore, conditional GAN (CGAN) introduces artificial control information on the basis of GAN, which is more practical many specific fields, though it mostly domain transfer. Researchers have proposed numerous methods to tackle diverse tasks by employing CGAN. It now timely and also critical point review these...

This paper focuses on the study of evolutionary mechanism governing temperature field geothermal reservoir under low-temperature tailwater reinjection conditions, which is crucial for sustainable energy management. With advancing exploitation resources deepens, precise understanding this becomes paramount devising effective strategies, optimizing utilization, and bolstering economic viability development. The article presents a comprehensive review evolution across diverse heterogeneous...

Aiming to solve the problems of poor dynamic response characteristics and weak anti-jamming capability conventional proportional–integral–derivative (PID) controlled pump-motor servo system (PMSS) under actual working environment, this study created a brand new hybrid grey wolf optimization (GWO) particle swarm (PSO) algorithm determine best parameters PID controller for PMSS speed control make achieve constant control. We developed GWOPSO-PID compared it with controller, GWO-PID, PSO-PID,...

In the existing reinforcement learning (RL)-based neural architecture search (NAS) methods for a generative adversarial network (GAN), both generator and discriminator are usually treated as objects. this article, we take different perspective to propose an approach by treating objective judge evaluate performance of architecture. Consequently, can convert NAS problem GAN-style problem, similar using controller generate sequential data via in sequence GAN, except that our generates...

A discrete optimization method using genetic algorithms is developed for the of ship structures. In this method, constrained minimization problem first transformed into an unconstrained one by a penalty term depending on degree constraint violation. Since search procedure algorithm done based evaluation fitness function, further converted maximization function. The design variables are coded binary string finite length. from generation to carried out simple with operators reproduction,...

In recent years, reinforcement learning (RL) and gradient optimization were applied with neural architecture search algorithms in the generative adversarial network (GAN) to achieve their state-of-the-art (SOTA) performance. However, existing RL-based methods utilized calculation of inception score or Fréchet distance as reward value guide controller, which actually wasted much searching time. order improve efficiency without degradation performance, this article proposes recycling...

In order to detect image spam effectively, we propose a method based on K-labels propagation model (KLPM) in this paper. Specifically, the speeded up robust features (SURF) of each are extracted firstly. Then standardize image, an improved means clustering algorithm is used cluster these and get information M centers. Finally, after being labeled, all testing images classified into or ham via KLPM, which K-nearest neighbor (KNN) graph label propagating process. Experiments show that...

A complete description of trust relationship is key to construct a high precision model. But most existing models not only miss the negative information and hesitation trust, but also ignore discordance between text comments ratings. To solve problems, dynamic evaluation model based on affect intensity proposed. In model, sentimental polarities are calculated from words in comments. The corresponding relation evaluations property emotional vector can be described. Time series weights...

In this paper, the Genetic Algorithm is used to optimize engine room structure under static and dynamic constraints. A penalty function method proposed handle complicated constraint conditions which include stress, natural frequency, allowable magnitude of design variables. The new operator called objective elitism as well fitness introduced improve efficiency method. Varied probability mutation convergence algorithm without risk premature a poor solution. To discrete problems, we serial...

针对实体恶意推荐问题,提出了一种角色分离的信任评估模型(RSTrust).模型将实体在信任评估中承担的角色分为交易角色和推荐角色两类,分别用交易信任度和推荐信任度来描述其可信性,区分不同角色对实体不同信任度的影响;在计算实体全局信任度时,RSTrust 将推荐者的全局推荐信任度作为其推荐证据的可信权重,消除恶意推荐对全局信任度计算的干扰.分析和仿真结果表明,模型具有良好的抗恶意推荐能力和收敛性.;To address the recommendation problems from malicious entities, a role separation based trust evaluation model (RSTrust) is proposed in this paper. In RSTrust, roles which entities act during are classified into transaction and roles. Trust on entity therefore described as according to their...

In order to solve the education problems caused by teachers and students’ unavoidable absence in school during COVID-19 pandemic, a series of online activities were carried out Nanjing University Posts Telecommunication early March. To explore students teachers’ degree satisfaction with distance education, this paper investigates multiple dimensions such as teachers, regional living standard, educational resources negative factors that reduce satisfaction, etc. Furthermore, attitude toward...

Abstract Present the method that using ACFM to test spherical tank which had appeared differential settlement. First introduced principle and system composition of ACFM, by which, pillar fillet weld LPG been tested, then quantitative evaluation cracks get out. After analyzed deeply, reliability when testing proved, in additionally, causes how come out given, too.

Taking advantage of the superiority valve controlled electro-hydraulic excitation system with wide vibration frequency and adjustable amplitude, a rotary is proposed. The structural composition principle alternating flow distribution due to are formulated. mathematical model derived simulation constructed based on Matlab/Simulink, accuracy verified by prototype experiments. In accordance this model, influence law main parameters hydraulic cylinder characteristics indicated. results reveal...

In practical design of ship structures, it is obvious that some structural elements, for instance, longitudinals, must be chosen from a given set standard discrete values being available commercially. Furthermore, integer parameters such as the number transverse web frames in midship segment should considered variables to optimized. view those facts, we investigate especially how deal with above kinds this paper. A cargo large hatch opening taken numerical example. The construction cost hold...