PSO和Cholesky分解的KELM的基因表达数据分类

核极限学习机(KELM)可使低维空间中线性不可分的数据变得线性可分,增加了ELM算法的鲁棒性,但KELM算法的输入权值参数采用随机初始化,容易导致算法不稳定.为此,本研究提出用粒子群优化算法对KELM中的权值初始参数进行优化、设定,以得到优化的分类器PSO-KELM.由于该算法输出权值求解采用传统的矩阵求逆运算,导致计算复杂,因此再对KELM的输出权值采用Cholesky分解进行优化.经一些标准基因数据集的实验表明,提出的PSO-KELM算法与已有的ELM、KELM、PSO-ELM相比分类精度更高,适用于基因表达数据分类.     

面向基因数据分类的AGA-ELM算法研究

对基因表达数据进行分类时,超限学习机(ELM)算法具有学习效率高、泛化能力强、分类精度高的优点.为了解决超限学习机算法受输入权值矩阵和隐含层偏差随机初始化的影响,本文利用自适应遗传算法(AGA)具有良好的全局搜索效果对超限学习机的输入权值矩阵和隐含层偏差进行优化,提出了基于自适应遗传算法优化超限学习机(AGA-ELM)的分类算法.通过实验表明,该算法与已有的ELM、GA-ELM以及SVM算法相比,分类精度更高,可用于基因数据分类.     

一种新的矢量量化初始码书算法

矢量量化的核心问题是初始码书设计。针对已有的随机法和分裂法等初始码书算法存在无效码矢多、运算量大等缺点,在训练矢量集随机抽取法的基础上提出一种新的初始码书算法即方差排序法,并应用到基于自组织特征映射(SOM)的矢量量化(VQ)中。实验结果表明,当码书尺寸大于256时,均值排序法的峰值信噪比(PSNR)比训练矢量集随机抽取法、分离平均法高,而且随码书尺寸的增加,峰值信噪比的改善程度大。     

人工神经网络在建筑物物理力学参数识别中的应用

基于神经网络的BP算法,建立了识别建筑物物理力学参数的数值方法.在经典的BP算法中,网络中的权值的确定是将权值的计算修正过程描述为网络的训练过程,其中迭代步长由经验选定,常常造成收敛速度慢,并且收敛速度与初始权值的选择有关以及引起振荡等问题.在网络的训练过程中采用改进的BP算法,通过对学习算子的优化搜索,大大提高了网络的收敛速度,解决了BP算法迭代过程中目标函数的震荡问题.数值计算表明,所提出的改进的BP算法进行建筑结构物理力学参数识别的收敛速度和识别精度都得到了提高.     

基于信息熵改进的粒子群算法在桁架结构

提出了信息熵改进的粒子群优化算法用于解决有应力约束、位移约束的桁架结构杆件截面尺寸优化设计问题.首先介绍了信息熵基本理论和基本粒子群优化算法理论,然后对粒子群优化算法作了合理的参数设置,并将信息熵引入粒子群优化算法的适应函数和停机判别准则中.最后对2个经典的优化问题进行求解并与其他算法进行了比较.数据结果表明信息熵改进后的粒子群优化算法在桁架结构优化设计中优于其他同类算法.     

SOM—SVM模型在煤与瓦斯突出预测中的应用

将自组织特征映射网络和支持向量机进行优选组合,建立煤与瓦斯突出危险性预测的SOM—SVM模型,充分利用非监督学习算法SOM的数据压缩、特征抽取的功能特性对训练样本进行压缩去噪处理,为有导师学习算法SVM提供高质量的有标记样本,进而发挥SVM分类精度高的特性,同时提高其分类速率。通过现场实测数据进行煤与瓦斯突出危险性预测,结果表明：两种算法的结合对煤与瓦斯突出危险性预测是有效的,它与传统的预测方法相比,分类速度更快,容错能力更强,预测精度更高。     

基于Moore-Penrose逆矩阵的选择性集成

本文提出了一种基于Moore-Penrose逆矩阵的新型选择性集成学习算法.先独立训练出一批个体学习器并为每个学习器指定一个初始权值,然后应用基于Moore-Penrose逆矩阵的算法对这些权值进行优化,最后选择权值较大的个体学习器进行最终集成.本文提出的选择性集成学习算法方法简单、易于实现,执行效率高.对8个真实数据集的实验表明,该集成学习算法相对于一般的集成学习算法,可以采用更少的学习器而获得更高的泛化能力.     

基于分类矢量量化器的超声检测数据压缩方法

研究了基于分类矢量量化技术的超声数据压缩算法,提出了基于峰值数的超声数据分类方法以及基于峰值距离失真测度的码字搜索算法。本文提出的算法充分利用了超声信号的特征,能够很好地保存超声信号的峰值信息。实验表明,相对于普通矢量量化器,本文提出的分类矢量量化器能够在保持相同压缩比的情况下提高重构信号的信噪比。将此矢量量化器与霍夫曼编码相结合,其压缩比可达1：50,高于基于小波变换的压缩算法。     

基于IGOA-RBF的矿用风压传感器温度补偿研究

针对矿用风压传感器受环境温度影响导致风压测量不准确的问题,提出基于改进蝗虫算法优化径向基神经网络(IGOA-RBF)的误差补偿方法,来消除环境温度影响所带来的零点漂移和灵敏度漂移.首先建立基于RBF神经网络的温度补偿模型,利用蝗虫算法(GOA)对RBF的网络初始权值、激活函数的数据中心及扩展常数进行优化,来提高模型的补...     

转子故障数据集降维的CKLPMDP算法研究

针对传统的数据降维方法难以兼顾局部流形结构和多流形判别结构学习的问题,提出一种相关熵测度核局部保持多流形判别投影算法(correntropy kernel locality preserving multi-manifold discriminant projection, CKLPMDP)的转子故障数据集降维方法。该方法的显著特点是采用相关熵测度监督近邻图的构建,首先将数据集映射到高维核空间,然后在核空间中综合考虑数据集的局部流形结构和多流形判别结构信息,提取出最优表征故障数据集的低维敏感特征矢量,采用三维图直观地显示出低维分类效果,并以低维敏感特征矢量输入K近邻分类器(K-nearest neighbor, KNN)中的辨识率和聚类分析中类间距S_b、类内距S_w作为衡量降维效果的指标。通过双跨转子实验台的振动信号数据集进行验证,与其他几种典型特征提取方法对比,该方法能更有效地提取出局部流形和多流形判别信息,在转子故障辨识中表现出更好的分类性能。     

An Improved Unsupervised Image Segmentation Method Based on Multi-Objective Particle Swarm Optimization Clustering Algorithm

Most image segmentation methods based on clustering algorithms use single-objective function to implement image segmentation. To avoid the defect, this paper proposes a new image segmentation method based on a multi-objective particle swarm optimization (PSO) clustering algorithm. This unsupervised algorithm not only offers a new similarity computing approach based on electromagnetic forces, but also obtains the proper number of clusters which is determined by scale-space theory. It is experimentally demonstrated that the applicability and effectiveness of the proposed multi-objective PSO clustering algorithm.     

A New Model Using Multiple Feature Clustering and Neural Networks for Forecasting Hourly PM2.5 Concentrations,and Its Applications in China

Particulate matter with an aerodynamic diameter no greater than 2.5 μm (PM_2.5) concentration forecasting is desirable for air pollution early warning. This study proposes an improved hybrid model, named multi-feature clustering decomposition (MCD)–echo state network (ESN)–particle swarm optimization (PSO), for multi-step PM_2.5 concentration forecasting. The proposed model includes decomposition and optimized forecasting components. In the decomposition component, an MCD method consisting of rough sets attribute reduction (RSAR), k-means clustering (KC), and the empirical wavelet transform (EWT) is proposed for feature selection and data classification. Within the MCD, the RSAR algorithm is adopted to select significant air pollutant variables, which are then clustered by the KC algorithm. The clustered results of the PM_2.5 concentration series are decomposed into several sublayers by the EWT algorithm. In the optimized forecasting component, an ESN-based predictor is built for each decomposed sublayer to complete the multi-step forecasting computation. The PSO algorithm is utilized to optimize the initial parameters of the ESN-based predictor. Real PM_2.5 concentration data from four cities located in different zones in China are utilized to verify the effectiveness of the proposed model. The experimental results indicate that the proposed forecasting model is suitable for the multi-step high-precision forecasting of PM_2.5 concentrations and has better performance than the benchmark models.     

基于粒子群优化聚类的汽轮机组振动故障诊断

针对模糊C-均值聚类算法(FCM)容易陷入局部极值和对初始值敏感的不足,提出了一种新的模糊聚类算法(PFCM),新算法利用粒子群优化算法(PSO)全局寻优、快速收敛的特点,代替了FCM算法的基于梯度下降的迭代过程,使算法具有很强的全局搜索能力,很大程度上避免了FCM算法易陷入局部极值的缺陷,同时也降低了FCM算法对初始值的敏感度。将该算法应用于汽轮机组振动故障诊断中,与电厂运行实际故障状态对照,仿真结果表明该算法提高了故障诊断的正确率。为汽轮机振动故障诊断方法的研究提供了一种新的思路。     

Bayesian regularisation neural network based on artificial intelligence optimisation

Stock prediction is generally considered to be challenging and known for its high noise and strong nonlinearities in financial time series analysis. However, current forecasting models ignore the importance of model parameter optimisation and the use of recent data. In this article, a novel forecasting approach with a Bayesian-regularised artificial neural networks (BR-ANN) was proposed. The weight of the proposed model (BR-ANN) is determined by the particle swarm optimisation (PSO) algorithm. Daily market prices and financial technical indicators are utilised as inputs to predict the one day future closing price of the Shanghai (in China) composite index. The Bayesian-regularised network uses a probabilistic nature for the network weights and can reduce the potential for over-fitting and over-training. Our empirical study and the results of our K-line theory analysis indicate that PSO is determined to be an effective algorithm to optimise the parameters of the Bayesian neural network compared with other well-known prediction algorithms. In particular, the PSO model is more reliable than the simple Bayesian regularisation neural network near the local maximum value.     

Automatic clustering for generalised cell formation using a hybrid particle swarm optimisation

This paper considers the cell formation (CF) problem in which parts have alternative process routings and the number of machine cells is not known a priori. Very few studies address these two practical issues at the same time. This paper proposes an automatic clustering approach based on a hybrid particle swarm optimisation (PSO) algorithm that can automatically evolve the number and cluster centres of machine cells for a generalised CF problem. In the proposed approach, a solution representation, comprising an integer number and a set of real numbers, is adopted to encode the number of cells and machine cluster centres, respectively. Besides, a discrete PSO algorithm is utilised to search for the number of machine cells, and a continuous PSO algorithm is employed to perform machine clustering. Effectiveness of the proposed approach has been demonstrated for test problems selected from the literature and those generated in this study. The experimental results indicate that the proposed approach is capable of solving the generalised machine CF problem without predetermination of the number of cells.     

Guided particle swarm optimization method to solve general nonlinear optimization problems

The development of hybrid algorithms is becoming an important topic in the global optimization research area. This article proposes a new technique in hybridizing the particle swarm optimization (PSO) algorithm and the Nelder–Mead (NM) simplex search algorithm to solve general nonlinear unconstrained optimization problems. Unlike traditional hybrid methods, the proposed method hybridizes the NM algorithm inside the PSO to improve the velocities and positions of the particles iteratively. The new hybridization considers the PSO algorithm and NM algorithm as one heuristic, not in a sequential or hierarchical manner. The NM algorithm is applied to improve the initial random solution of the PSO algorithm and iteratively in every step to improve the overall performance of the method. The performance of the proposed method was tested over 20 optimization test functions with varying dimensions. Comprehensive comparisons with other methods in the literature indicate that the proposed solution method is promising and competitive.     

基于聚类的粒子群优化算法在MES调度中的应用

针对传统调度算法寻优效率低的弱点,从MES功能出发,将其调度功能单独抽出,提出了基于聚类的粒子群优化算法,将聚类用于粒子群搜索空间的改进。仿真结果表明了该算法的有效性。     

文本挖掘中运用自组织特征映射算法分析中国人类工效学研究状况

文本挖掘是抽取有效、新颖、有用、可理解的、散布在文本文件中的有价值知识,并且利用这些知识更好地组织信息的过程。利用文本挖掘中的自组织特征映射(SOM)算法,对中国《人类工效学》期刊数据库的大量文档进行聚类分析,得到当前国内人类工效学研究领域里的主要研究类别、趋势,然后将聚类结果与国际人类工效学协会(IEA)公布的研究领域进行对比分析。     

Unsupervised learning-based clustering approach for smart identification of pathologies and segmentation of tissues in brain magnetic resonance imaging

Human-made/developed algorithms provide automatic identification and segmentation of the tissues, lesions and tumor regions available in brain magnetic resonance scan images, which invocates predicaments such as high computational cost and low accuracy rate. Such hassles are reconciled with the utilization of an unsupervised approach in combination with clustering techniques. Initially, static features are chosen from the input image, which is fed to the self-organizing map (SOM), where the algorithm employs the dimensionality reduction of input images. Consecutively, the reduced SOM prototype of data is clustered by the modified fuzzy K-means (MFKM) algorithm. The MFKM algorithm can be modified in terms of membership variables because it operates with spatial information and converges quickly, and this would be of greater benefit to radiologists as they reduce the wrong predictions and voluminous time that normally occur owing to human involvement. The proposed algorithm provides 98.77% sensitivity and 97.5% specificity, which are better than any other traditional algorithms mentioned in this article.     

Using artificial neural networks for mapping of scienceand technology: A multi-self-organizing-maps approach

We argue in favour of artificial neural networks for exploratory data analysis, clustering andmapping. We propose the Kohonen self-organizing map (SOM) for clustering and mappingaccording to a multi-maps extension. It is consequently called Multi-SOM. Firstly the KohonenSOM algorithm is presented. Then the following improvements are detailed: the way of namingthe clusters, the map division into logical areas, and the map generalization mechanism. Themulti-map display founded on the inter-maps communication mechanism is exposed, and thenotion of the viewpoint is introduced. The interest of Multi-SOM is presented for visualization,exploration or browsing, and moreover for scientific and technical information analysis. A casestudy in patent analysis on transgenic plants illustrates the use of the Multi-SOM. We also showthat the inter-map communication mechanism provides support for watching the plants on whichpatented genetic technology works. It is the first map. The other four related maps provideinformation about the plant parts that are concerned, the target pathology, the transgenictechniques used for making these plants resistant, and finally the firms involved in geneticengineering and patenting. A method of analysis is also proposed in the use of this computerbasedmulti-maps environment. Finally, we discuss some critical remarks about the proposedapproach at its current state. And we conclude about the advantages that it provides for aknowledge-oriented watching analysis on science and technology. In relation with this remark weintroduce in conclusion the notion of knowledge indicators.