To automatically mine structured semantic topics from text, neural topic modeling has arisen and made some progress. However, most existing work focuses on designing a mechanism to enhance topic coherence but sacrificing the diversity of the extracted topics. To address this limitation, we propose the first neural-based topic modeling approach purely based on mutual information maximization, called the mutual information topic (MIT) model, in this article. The proposed MIT significantly improves topic diversity by maximizing the mutual information between word distribution and topic distribution. Meanwhile, MIT also utilizes Dirichlet prior in latent topic space to ensure the quality of mined topics. The experimental results on three publicly benchmark text corpora show that MIT could extract topics with higher coherence values (considering four topic coherence metrics) than competitive approaches and has a significant improvement on topic diversity metric. Besides, our experiments prove that the proposed MIT converges faster and more stable than adversarial-neural topic models.

Summary The wireless network resource allocation is a NP‐hard combinatorial optimization problem. This paper proposes a new optimization method using the manta ray foraging optimization (MRFO) based on spiral modeling to solve the wireless network resource problem. In order to reduce the computational complexity and ensure the optimal performance of the allocation scheme, a MRFO algorithm based on spiral modeling and mutation strategy is proposed. In the first stage, spiral modeling is introduced to narrow the exploration area, while the mutation strategy of the genetic algorithm enhances the ability of the algorithm to jump out of the local optimal. In the second stage, a binary MRFO algorithm for using different transfer functions is proposed to solve the mixed integer programming problem. The experimental results show that IMRFO and BIMRFO have high comprehensive performance advantages, achieve better effects than the other optimization algorithms, which lead to faster convergence, faster accuracy, and lower complexity.

Summary In this article, a new quantum social spider optimization (QSSO) algorithm is proposed. In the QSSO algorithm, we introduce an encoding approach based on bits described on social spider optimization (SSO) and serves as the evolution method of the population space. For the encoding of individuals, the probability amplitude expression of quantum bit is applied to describe the position of individuals, by which one individual's position can be expressed as the superposition of multistates. In such a way, the population diversity and the global searching capability of the SSO algorithm are enhanced. The QSSO algorithm was used to optimize the hydropower station dispatch, and the calculation results show that QSSO algorithm has fast convergence, small number of tuning parameters, high calculation accuracy, stability, simple, and is easy to be implemented with strong global search capability.

Abstract Salp swarm algorithm (SSA) is one of the efficient recent meta‐heuristic optimization algorithms, where it has been successfully utilized in a wide range of optimization problems in different fields. In the research process, it is found that it is very difficult to maintain the balance between the exploration and exploitation capabilities of a certain algorithm. Therefore, one of the main purposes of this article is to provide an algorithm that can intelligently balance between exploration and exploitation, so that it can balance exploration and exploitation capabilities. Later, in the research process, it was found that the sine and cosine function and the salp foraging trajectory have a high mathematical similarity, which greatly improves the optimization ability of the algorithm. In addition, the variable neighbourhood strategy can appropriately expand the optimization range of the algorithm. So in this paper, a novel golden sine cosine salp swarm algorithm with variable neighbourhood search scheme (GSCSSA‐VNS) is proposed, the another objective of proposing this algorithm is as a new optimization method for shape matching. As a relatively new branch, atomic potential matching (APM) model is inspired by potential field attractions. Compared to the conventional edge potential function (EPF) model, APM has been verified to be less sensitive to intricate backgrounds in the test image and far more cost effective in the computation process. Experimental results of four realistic examples show that GSCSSA‐VNS is able to provide very competitive results and outperforms the other algorithms.

Summary Dragonfly algorithm (DA) is a new swarm intelligence optimization algorithm based on the static and dynamic swarm behavior of dragonflies. The algorithm has the characteristics of simple structure, strong search ability, easy implementation, and strong robustness. However, the DA algorithm itself also has insufficient solution accuracy and slow convergence speed. The Wind Driven Optimization algorithm (WDO) has the characteristics of fast convergence speed and strong global search capability. So as to improve the optimization performance of the DA algorithm and avoid premature convergence, the speed of the WDO is introduced into the later calculation of the algorithm iteration, which speeds up the convergence speed of the global optimal solution. This paper proposes a dragonfly algorithm based on wind driven (WDDA), that is to reduce the blindness of the dragonfly algorithm search, improve the solution accuracy and convergence speed, to improve the overall optimization performance of the algorithm. The 23 benchmark test functions and one engineering example for optimization and comparison experiments. The experimental results show that WDDA algorithm has better performance in function optimization.

Networks.He is working on the application of multi-objective and robust meta-heuristic optimization techniques.

To overcome the poor population diversity and slow convergence rate of grey wolf optimizer (GWO), this paper introduces the elite opposition-based learning strategy and simplex method into GWO, and proposes a hybrid grey optimizer using elite opposition (EOGWO). The diversity of grey wolf population is increased and exploration ability is improved. The experiment results of 13 standard benchmark functions indicate that the proposed algorithm has strong global and local search ability, quick convergence rate and high accuracy. EOGWO is also effective and feasible in both low-dimensional and high-dimensional case. Compared to particle swarm optimization with chaotic search (CLSPSO), gravitational search algorithm (GSA), flower pollination algorithm (FPA), cuckoo search (CS) and bat algorithm (BA), the proposed algorithm shows a better optimization performance and robustness.

Path planning of Unmanned Undersea Vehicle (UUV) is a rather complicated global optimum problem which is about seeking a superior sailing route considering the different kinds of constrains under complex combat field environment. Flower pollination algorithm (FPA) is a new optimization method motivated by flower pollination behavior. In this paper, a variant of FPA is proposed to solve the UUV path planning problem in two-dimensional (2D) and three-dimensional (3D) space. Optimization strategies of particle swarm optimization are applied to the local search process of IFPA to enhance its search ability. In the progress of iteration of this improved algorithm, a dimension by dimension based update and evaluation strategy on solutions is used. This new approach can accelerate the global convergence speed while preserving the strong robustness of standard FPA. The realization procedure for this improved flower pollination algorithm is also presented. To prove the performance of this proposed method, it is compared with nine population-based algorithms. The experiment result shows that the proposed approach is more effective and feasible in UUV path planning in 2D and 3D space.