Randomised Local Search Algorithm for the Clustering Problem

We consider clustering as a combinatorial optimisation problem. Local search provides a simple and effective approach to many other combinatorial optimisation problems. It is therefore surprising how seldom it has been applied to the clustering problem. Instead, the best clustering results have been obtained by more complex techniques such as tabu search and genetic algorithms at the cost of high run time. We introduce a new randomised local search algorithm for the clustering problem. The algorithm is easy to implement, sufficiently fast, and competitive with the best clustering methods. The ease of implementation makes it possible to tailor the algorithm for various clustering applications with different distance metrics and evaluation criteria.     

Differential evolution and differential ant-stigmergy on dynamic optimisation problems

Many real-world optimisation problems are of dynamic nature, requiring an optimisation algorithm which is able to continuously track a changing optimum over time. To achieve this, we propose two population-based algorithms for solving dynamic optimisation problems (DOPs) with continuous variables: the self-adaptive differential evolution algorithm (jDE) and the differential ant-stigmergy algorithm (DASA). The performances of the jDE and the DASA are evaluated on the set of well-known benchmark problems provided for the special session on Evolutionary Computation in Dynamic and Uncertain Environments. We analyse the results for five algorithms presented by using the non-parametric statistical test procedure. The two proposed algorithms show a consistently superior performance over other recently proposed methods. The results show that both algorithms are appropriate candidates for DOPs.     

Transfer learning from synthetic labels for histopathological images classification

This study introduces a new strategy that combines unsupervised learning (clustering) and transfer learning. Clustering methods are employed to generate synthetic labels for the source dataset (ICAR-2018). The generated dataset is then used for transfer learning to other histopathological datasets (KimiaPath960, CRC, Biomaging??2015, Breakhis, and Lymphoma). The comparative study based on two clustering algorithms (K-means and multi-objective clustering stream) demonstrates the efficiency of MOC-Stream. The generated synthetic histopathological dataset by this clustering algorithm outperformed the original labeled dataset and the imageNet models in transfer learning.

     

基于样本空间分布密度的初始聚类中心优化K-均值算法*

针对传统K-均值聚类算法对初始聚类中心敏感、现有初始聚类中心优化算法缺乏客观性,提出一种基于样本空间分布密度的初始聚类中心优化K-均值算法。该算法利用数据集样本的空间分布信息定义数据对象的密度,并根据整个数据集的空间信息定义了数据对象的邻域;在此基础上选择位于数据集样本密集区且相距较远的数据对象作为初始聚类中心,实现K-均值聚类。UCI机器学习数据库数据集以及随机生成的带有噪声点的人工模拟数据集的实验测试证明,本算法不仅具有很好的聚类效果,而且运行时间短,对噪声数据有很强的抗干扰性能。基于样本空间分布密度的初始聚类中心优化K-均值算法优于传统K-均值聚类算法和已有的相关K-均值初始中心优化算法。     

基于MapReduce的分布式近邻传播聚类算法

随着信息技术迅速发展,数据规模急剧增长,大规模数据处理非常具有挑战性.许多并行算法已被提出,如基于MapReduce的分布式K平均聚类算法、分布式谱聚类算法等.近邻传播(affinity propagation,AP)聚类能克服K平均聚类算法的局限性,但是处理海量数据性能不高.为有效实现海量数据聚类,提出基于MapReduce的分布式近邻传播聚类算法——DisAP.该算法先将数据点随机划分为规模相近的子集,并行地用AP聚类算法稀疏化各子集,然后融合各子集稀疏化后的数据再次进行AP聚类,由此产生的聚类代表作为所有数据点的聚类中心.在人工合成数据、人脸图像数据、IRIS数据以及大规模数据集上的实验表明:DisAP算法对数据规模有很好的适应性,在保持AP聚类效果的同时可有效缩减聚类时间.     

GSA-LA: gravitational search algorithm based on learning automata

ABSTRACT

Regardless of the performance of gravitational search algorithm (GSA), it is nearly incapable of avoiding local optima in high-dimension problems. To improve the accuracy of GSA, it is necessary to fine tune its parameters. This study introduces a gravitational search algorithm based on learning automata (GSA-LA) for optimisation of continuous problems. Gravitational constant G(t) is a significant parameter that is used to adjust the accuracy of the search. In this work, learning capability is utilised to select G(t) based on spontaneous reactions. To measure the performance of the introduced algorithm, numerical analysis is conducted on several well-designed test functions, and the results are compared with the original GSA and other evolutionary-based algorithms. Simulation results demonstrate that the learning automata-based gravitational search algorithm is more efficient in finding optimum solutions and outperforms the existing algorithms.     

基于核的K-均值聚类

将核学习方法的思想应用于K-均值聚类中，提出了一种核K-均值聚类算法，算法的主要思想是：首先将原空间中待聚类的样本经过一个非线性映射，映射到一个高维的核空间中，突出各类样本之间的特征差异，然后在这个核空间中进行K-均值聚类。同时还将一种新的核函数应用于核K-均值聚类中以提高算法的速度。为了验证算法的有效性，分别利用人工和实际数据进行K-均值聚类和核K-均值聚类，实验结果显示对于一些特殊的类分布数据，核K-均值聚类比K-均值聚类具有更好的聚类效果。     

Particle swarm optimisation in designing parameters of manufacturing processes: A review (2008–2018)

The evolutionary optimisation algorithms appeared as an effective alternative to conventional statistical methods that have certain limitations in optimising complex manufacturing processes. Considering works published in the last decade, this paper presents an analysis of the particle swarm optimisation (PSO) implementation in designing parameters of heterogeneous manufacturing processes, both conventional and emerging, new processes. The literature review and analysis was structured according to the complexity of the optimisation problem (single response and multiresponse problems), and the development of an objective function for PSO. The tuning of the PSO algorithm-specific parameters was analysed in detail. The PSO algorithm performance was benchmarked with the results of other methods, including evolutionary algorithms, in designing process parameters. The concerns in applying PSO for multiresponse manufacturing problems were highlighted, and recommendations for future research were drawn. Such a comprehensive review on the PSO application in optimising manufacturing processes, including the detailed discussion on the algorithm characteristics and benchmark with other optimisation procedures, has not been pursued so far. Therefore, this review analysis provides hands on information for researchers and engineers at one place, and it is believed that the findings could serve as a basis for the future research and implementation directions.     

Handling many-objective optimisation problems with R2 indicator and decomposition-based particle swarm optimiser

An R2 indicator-based multi-objective particle swarm optimiser (R2-MOPSO) can obtain well-convergence and well-distributed solutions while solving two and three objectives optimisation problems. However, R2-MOPSO faces difficulty to tackle many-objective optimisation problems because balancing convergence and diversity is a key issue in high-dimensional objective space. In order to address this issue, this paper proposes a novel algorithm, named R2-MaPSO, which combines the R2 indicator and decomposition-based archiving pruning strategy into particle swarm optimiser for many-objective optimisation problems. The innovations of the proposed algorithm mainly contains three crucial factors: (1) A bi-level archiving maintenance approach based on the R2 indicator and objective space decomposition strategy is designed to balance convergence and diversity. (2) The global-best leader selection is based on the R2 indicator and the personal-best leader selection is based on the Pareto dominance. Meanwhile, the objective space decomposition leader selection adopts the feedback information from the bi-level archive. (3) A new velocity updated method is modified to enhance the exploration and exploitation ability. In addition, an elitist learning strategy and a smart Gaussian learning strategy are embedded into R2-MaPSO to help the algorithm jump out of the local optimal front. The performance of the proposed algorithm is validated and compared with some algorithms on a number of unconstraint benchmark problems, i.e. DTLZ1-DTLZ4, WFG test suites from 3 to 15 objectives. Experimental results have demonstrated a better performance of the proposed algorithm compared with several multi-objective particle swarm optimisers and multi-objective evolutionary algorithms for many-objective optimisation problems.     

云模式事件混沌关联特征提取的物联网大数据聚类算法

目前的聚类方法单纯从某个角度研究数据聚类问题,对基于云模式的混沌的物联网大数据聚类的考虑不足,聚类质量不高。为实现敏捷、智能、平稳的物联网大数据聚类,基于开展物联网事件的云模式通用描述模型、物联网事件混沌关联特征的云模式通用解析模型、基于云模式的物联网事件混沌关联特征提取算法、基于云模式混沌关联特征的物联网大数据关联挖掘研究,改进分解奇异值算法、网格耦合聚类算法、K-means算法、决策树学习法、分析主成分法、分层合并法等算法和分布概率函数,设计了一种基于事件混沌关联特征、敏捷、智能、平稳的物联网大数据聚类算法。最后,开展实验验证,并与传统算法进行性能对比分析。实验结果表明,相比传统算法,该算法聚类时间短、误差小,且敏捷性、智能性、动态演化性和平稳性高。因此,该算法实现了基于云模式的具有混沌关联特征的物联网事件大数据的有效聚类,具有较高的应用价值。     

一种改进的求解聚类问题的差分进化算法

针对K-means算法依赖于初始聚类中心和易陷入局部最优解的缺陷,提出一种改进的求解聚类问题的差分进化算法。将改进的差分进化算法和K-means迭代相结合,使算法对初始聚类中心的敏感性和陷入局部最优解的可能性降低,提高了算法的稳定性。通过将反向学习技术引入到框架中来指导搜索新的空间,提高了算法的全局寻优能力。为了提高算法效率,根据聚类问题编码的特点设计了一种整理算子来消除冗余以及调整了差分进化算法的种群更新策略。最后在迭代过程中不断引入随机个体,增强了种群的多样性。与K-means和几个进化聚类算法进行比较,实验结果表明,该算法不仅能有效抑制早熟收敛,而且具有较强的稳定性,较好的聚类效果。     

一种改进的多视图聚类集成算法

近年来,针对大数据的数据挖掘技术和机器学习算法研究变得日趋重要。在聚类领域,随着多视图数据的大量出现,多视图聚类已经成为了一类重要的聚类方法。然而,大多数现有的多视图聚类算法受算法参数设置、数据样本等影响,具有聚类结果不稳定、参数需要反复调节等缺点。基于多视图K-means算法和聚类集成技术,提出了一种改进的多视图聚类集成算法,其提高了聚类的准确性、鲁棒性和稳定性。其次,由于单机环境下的多视图聚类算法难以对海量的数据进行处理,结合分布式处理技术,实现了一种分布式的多视图并行聚类算法。实验证明,并行算法在处理大数据时的时间效率有很大提升,适合于大数据环境下的多视图聚类分析。     

Multi-objective ACO algorithms to minimise the makespan and the total rejection cost on BPMs with arbitrary job weights

In this paper, we investigate the batch-scheduling problem with rejection on parallel machines with non-identical job sizes and arbitrary job-rejected weights. If a job is rejected, the corresponding penalty has to be paid. Our objective is to minimise the makespan of the processed jobs and the total rejection cost of the rejected jobs. Based on the selected multi-objective optimisation approaches, two problems, P1 and P2, are considered. In P1, the two objectives are linearly combined into one single objective. In P2, the two objectives are simultaneously minimised and the Pareto non-dominated solution set is to be found. Based on the ant colony optimisation (ACO), two algorithms, called LACO and PACO, are proposed to address the two problems, respectively. Two different objective-oriented pheromone matrices and heuristic information are designed. Additionally, a local optimisation algorithm is adopted to improve the solution quality. Finally, simulated experiments are conducted, and the comparative results verify the effectiveness and efficiency of the proposed algorithms, especially on large-scale instances.     

Text Clustering with Feature Selection by Using Statistical Data

Feature selection is an important method for improving the efficiency and accuracy of text categorization algorithms by removing redundant and irrelevant terms from the corpus. In this paper, we propose a new supervised feature selection method, named CHIR, which is based on the chi² statistic and new statistical data that can measure the positive term-category dependency. We also propose a new text clustering algorithm, named text clustering with feature selection (TCFS). TCFS can incorporate CHIR to identify relevant features (i.e., terms) iteratively, and the clustering becomes a learning process. We compared TCFS and the K-means clustering algorithm in combination with different feature selection methods for various real data sets. Our experimental results show that TCFS with CHIR has better clustering accuracy in terms of the F-measure and the purity.     

Parallel hyper heuristic algorithm based on reinforcement learning for the corridor allocation problem and parallel row ordering problem

Hyper heuristics is a relatively new optimisation algorithm. Numerous studies have reported that hyper heuristics are well applied in combinatorial optimisation problems. As a classic combinatorial optimisation problem, the row layout problem has not been publicly reported on applying hyper heuristics to its various sub-problems. To fill this gap, this study proposes a parallel hyper-heuristic approach based on reinforcement learning for corridor allocation problems and parallel row ordering problems. For the proposed algorithm, an outer layer parallel computing framework was constructed based on the encoding of the problem. The simulated annealing, tabu search, and variable neighbourhood algorithms were used in the algorithm as low-level heuristic operations, and Q-learning in reinforcement learning was used as a high-level strategy. A state space containing sequences and fitness values was designed. The algorithm performance was then evaluated for benchmark instances of the corridor allocation problem (37 groups) and parallel row ordering problem (80 groups). The results showed that, in most cases, the proposed algorithm provided a better solution than the best-known solutions in the literature. Finally, the meta-heuristic algorithm applied to three low-level heuristic operations is taken as three independent algorithms and compared with the proposed hyper-heuristic algorithm on four groups of parallel row ordering problem instances. The effectiveness of Q-learning in selection is illustrated by analysing the comparison results of the four algorithms and the number of calls of the three low-level heuristic operations in the proposed method.     

Seeker optimisation (human group optimisation) algorithm with chaos

Seeker optimisation algorithm (SOA), also referred to as human group metaheuristic optimisation algorithms form a very hot area of research, is an emerging population-based and gradient-free optimisation tool. It is inspired by searching behaviour of human beings in finding an optimal solution. The principal shortcoming of SOA is that it is easily trapped in local optima and consequently fails to achieve near-global solutions in complex optimisation problems. In an attempt to relieve this problem, in this article, chaos-based strategies are embedded into SOA. Five various chaotic-based SOA strategies with four different chaotic map functions are examined and the best strategy is chosen as the suitable chaotic scheme for SOA. The results of applying the proposed chaotic SOA to miscellaneous benchmark functions confirm that it provides accurate solutions. It surpasses basic SOA, genetic algorithm, gravitational search algorithm variant, cuckoo search optimisation algorithm, firefly swarm optimisation and harmony search the proposed chaos-based SOA is expected successfully solve complex engineering optimisation problems.     

Multi-Cohort Intelligence algorithm: an intra- and inter-group learning behaviour based socio-inspired optimisation methodology

ABSTRACT

A Multi-Cohort Intelligence (Multi-CI) metaheuristic algorithm in emerging socio-inspired optimisation domain is proposed. The algorithm implements intra-group and inter-group learning mechanisms. It focusses on the interaction amongst different cohorts. The performance of the algorithm is validated by solving 75 unconstrained test problems with dimensions up to 30. The solutions were comparing with several recent algorithms such as Particle Swarm Optimisation (PSO), Covariance Matrix Adaptation Evolution Strategy, Artificial Bee Colony, Self-Adaptive Differential Evolution Algorithm, Comprehensive Learning Particle Swarm Optimisation, Backtracking Search Optimisation Algorithm, and Ideology Algorithm. The Wilcoxon signed-rank test was carried out for the statistical analysis and verification of the performance. The proposed Multi-CI outperformed these algorithms in terms of the solution quality including objective function value and computational cost, i.e. computational time and functional evaluations. The prominent feature of the Multi-CI algorithm along with the limitations is discussed as well. In addition, an illustrative example is also solved and every detail is provided.     

A new effective operator for the hybrid algorithm for solving global optimisation problems

Hybrid algorithms have been recently used to solve complex single-objective optimisation problems. The ultimate goal is to find an optimised global solution by using these algorithms. Based on the existing algorithms (HP_CRO, PSO, RCCRO), this study proposes a new hybrid algorithm called MPC (Mean-PSO-CRO), which utilises a new Mean-Search Operator. By employing this new operator, the proposed algorithm improves the search ability on areas of the solution space that the other operators of previous algorithms do not explore. Specifically, the Mean-Search Operator helps find the better solutions in comparison with other algorithms. Moreover, the authors have proposed two parameters for balancing local and global search and between various types of local search, as well. In addition, three versions of this operator, which use different constraints, are introduced. The experimental results on 23 benchmark functions, which are used in previous works, show that our framework can find better optimal or close-to-optimal solutions with faster convergence speed for most of the benchmark functions, especially the high-dimensional functions. Thus, the proposed algorithm is more effective in solving single-objective optimisation problems than the other existing algorithms.     

用于求解对称旅行商问题的粒子群算法和蚂蚁算法的融合

近年来,基于仿生学的随机优化技术成为学术界研究的重点问题之一,并在许多领域得到应用。粒子群优化（PSO）算法和蚂蚁算法ACO（Ant Colong Optimization）是随机全局优化的两个重要方法。PSO算法初始收敛速度较快,但在接近最优解时,收敛速度较慢,而ACO正好相反。结合二者的优势,先利用粒子群算法,再结合蚂蚁算法,以对称旅行商问题为例进行了仿真实现。实验结果表明,先利用PSO算法进行初步求解,在利用蚂蚁算法进行精细求解,可以得到较好的效果。     

基于余弦距离选取初始簇中心的文本聚类研究

文本聚类是文本信息进行有效组织、摘要和导航的重要手段,其中基于余弦相似度的K-means算法是最重要且使用最广泛的文本聚类算法之一。针对基于余弦相似度的K-means算法改进方案设计困难,且众多优异的基于欧氏距离的K-means改进方法无法适用的问题,对余弦相似度与欧氏距离的关系进行探讨,得到标准向量前提下二者的转化公式,并在此基础上定义一种与欧氏距离意义相近关系紧密的余弦距离,使原有基于欧氏距离的K-means改进方法可通过余弦距离迁移到基于余弦相似度的K-means算法中。在此基础上理论推导出余弦K-means算法及其拓展算法的簇内中心点计算方法,并进一步改进了聚类初始簇中心的选取方案,形成新的文本聚类算法MCSKM++。通过实验验证,该算法在迭代次数减少、运行时间缩短的同时,聚类精度得到提高。