PSO-SOM分类判别研究及其应用

针对网络初始权矢量选取的不确定性问题,提出了粒子群优化-自组织映射(PSO-SOM)算法,利用PSO算法优化SOM网络的初始权矢量,进而进行分类.将提出的方法用于基因表达数据的分类判别中,使得SOM网络的误差平方和大大下降,提高了网络的分类精度,表明PSO-SOM算法用于数据的分类判别是切实有效的.     

A DNA-based algorithm for minimizing decision rules: a rough sets approach

Rough sets are often exploited for data reduction and classification. While they are conceptually appealing, the techniques used with rough sets can be computationally demanding. To address this obstacle, the objective of this study is to investigate the use of DNA molecules and associated techniques as an optimization vehicle to support algorithms of rough sets. In particular, we develop a DNA-based algorithm to derive decision rules of minimal length. This new approach can be of value when dealing with a large number of objects and their attributes, in which case the complexity of rough-sets-based methods is NP-hard. The proposed algorithm shows how the essential components involved in the minimization of decision rules in data processing can be realized.     

Predicting glass furnace output using statistical and neural computing methods

This paper describes the development of predictive models for glass production at a regional manufacturing company. The objectives of the models are to predict the actual batch tonnage produced per week from the glass furnace based on the planned production schedule. Four modelling methods were explored: (i) linear regression; (ii) nonlinear regression; (iii) artificial neural network using back-propagation; and (iv) radial basis function neural network. Using 175 cases of production schedule data and subsequent furnace output, the two neural network-based prediction models resulted in lower average absolute error and lower maximum absolute error than the linear or nonlinear regression models. Accurate neural network-based prediction models of furnace output will subsequently be used in the overall production planning system by utilizing estimates of furnace output to determine the necessary energy, raw material, repair and personnel requirements of the glass manufacturing facility.     

High-order statistical blind deconvolution of spectroscopic data with a Gauss-Newton algorithm

The spectroscopic data recorded by a dispersion spectrophotometer are usually degraded by the response function of the instrument. To improve the resolving power, double or triple cascade spectrophotometers and narrow slits have been employed, but the total flux of the radiation available decreases accordingly, resulting in a lower signal-to-noise ratio (SNR) and a longer measurement time. However, the spectral resolution can be improved by mathematically removing the effect of the instrument response function. A high-order statistical Gauss-Newton algorithm is proposed to blindly deconvolve the measured spectroscopic data. The true spectrum and the instrument response function are estimated simultaneously. Experiments on artificial and real measured spectroscopic data demonstrate the feasibility of this method.     

Robust integral image rectification framework using perspective transformation supported by statistical line segment clustering

In most integral image analysis and processing tasks, accurate knowledge of the internal image structure is required. In this paper we present a robust framework for the accurate rectification of perspectively distorted integral images based on multiple line segment detection. The use of multiple line segments increases the overall fault tolerance of our framework providing strong statistical support for the rectification process. The proposed framework is used for the automatic rectification, metric correction, and rotation of distorted integral images. The performance of our framework is assessed over a number of integral images with varying scene complexity and noise levels.     

Machine-component grouping in production flow analysis: an approach using a rank order clustering algorithm

Existing cluster analysis methods are reviewed and a new approach using a rank order clustering algorithm is described which is particularly relevant to the problem of machine-component group formation. A relaxation and regrouping procedure is developed whereby the basic rank order clustering method may be extended to the case where there are bottleneck machines.     

Estimating the drift time for processes subject to linear trend disturbance using fuzzy statistical clustering

Although control charts can notify the state of out-of-control in a process by generating a signal, the indication is usually followed by a considerable amount of delay. Identifying the real time of the change in a process would provide a starting point for further investigation of an assignable cause. This paper addresses the problem of detecting the change point in different processes when the quality characteristics drift steadily away from an in-control state. For this purpose, a fuzzy statistical clustering (FSC) method is used to estimate the drift time in different processes. Since the application of an FSC method requires both in- and out-of-control values of the process parameter, a linear regression model is utilised to estimate the trend rate and then calculate the out-of-control process parameter. Through extensive simulations, the performance of the proposed change point estimation method is analysed and compared with the most recent estimators for several control charts. The results demonstrate that the proposed method is more effective in detecting the drift time through a wide range of trend rates. Furthermore, it is shown that the proposed method offers a higher estimation precision compared to conventional statistical methods.     

Visual hierarchical clustering of supply chain using growing hierarchical self-organising map algorithm

The study identifies a need for efficient and robust visual clustering approach that can potentially deal with complex supply chain clustering problems. Based on the underlying philosophy of group technology, a growing hierarchical self-organising map algorithm (GHSOM) is proposed to identify a lower two-dimension visual clustering map that can effectively address supply chain clustering problems. The proposed approach provides optimal solutions by decomposing a large-sized supply chain problem into independent, small, manageable problems. It facilitates simple decision-making by exploring similar clusters that are represented by the neighbouring branches in the GHSOM map structure. Unlike other approaches in literature, the proposed approach can further attain good topological ordered representations of the various work order families, to be processed by clusters of supply units along with information on hierarchical sub-cell formation as identifiable from the visually navigable map. The proposed approach has been successfully applied on 16 benchmarked problems. The performance of GHSOM based on grouping efficacy measure outperformed the best results in literature.     

一种采用背景统计技术的视频对象分割算法

利用背景统计技术从累积的帧差信息中构建出完整、可靠的背景区域,并将其与当前帧相比较,得到初始对象分割掩膜;再对之进行后处理,以消除噪声影响和平滑对象边界,从而获得较好的对象分割掩膜,并提取出视频对象。该算法不需要预知运动对象的形状、数目等,就能较好地从静止背景中分离出目标,实验证明,它具有一定的实用性和鲁棒性。     

Control of enhanced Raman scattering using a DNA-based assembly process of dye-coded nanoparticles

Enhanced Raman scattering from metal surfaces has been investigated for over 30 years. Silver surfaces are known to produce a large effect, and this can be maximized by producing a roughened surface, which can be achieved by the aggregation of silver nanoparticles. However, an approach to control this aggregation, in particular through the interaction of biological molecules such as DNA, has not been reported. Here we show the selective turning on of the surface enhanced resonance Raman scattering effect on dye-coded, DNA-functionalized, silver nanoparticles through a target-dependent, sequence-specific DNA hybridization assembly that exploits the electromagnetic enhancement mechanism for the scattering. Dye-coded nanoparticles that do not undergo hybridization experience no enhancement and hence do not give surface enhanced resonance Raman scattering. This is due to the massive difference in enhancement from nanoparticle assemblies compared with individual nanoparticles. The electromagnetic enhancement is the dominant effect and, coupled with an understanding of the surface chemistry, allows surface enhanced resonance Raman scattering nanosensors to be designed based on a natural biological recognition process.     

A clustering algorithm for machine cell formation in group technology using minimum spanning trees

I address the machine cell part family formation problem in group technology. The minimum spanning tree (MST) For machines is constructed from which seeds to cluster components are generated. Seeds to cluster machines are obtained from component clusters. The process of alternate seed generation and clustering is continued until feasible solutions are obtained. Edges are removed from the MST to identify alternate starting seeds for clustering. The algorithm is tested with matrices available in the literature. The results compare favourably with existing methods.     

土木工程系统辨识统计方法的现状与展望

土木工程由于内在材料的离散性,结构体系的复杂性,外在环境因素的随机性,有必要引入统计方法来进行面向健康监测的损伤辨识。探讨了土木工程系统辨识的定义,分析了现有研究中与统计方法相关的Bayesian统计、随机有限元、统计模式辨识、损伤特征辨识的统计描述及其他等五类研究;根据土木工程系统辨识的特点与已有研究,展望了统计方法进一步的研究前景与方向。     

A fast permutation-based algorithm for block clustering

A stepwise divisive procedure for the clustering of numerical data recorded in matrix form into homogeneous groups is introduced. The methodology relates to those proposed by Hartigan (1972) and Duffy and Quiroz (1991). As the latter, the proposed methodology uses the permutation distribution of the data in a block as the reference distribution to make inferences about the presence of clustering structure. A local (within block) criteria and Bayesian sequential decision methodology are used to evaluate the significance of potential partitions of blocks, resulting in an algorithm which is faster than those considered by Duffy and Quiroz (1991). The class of possible clustering structures that our procedure can discover is also larger than those previously considered in the literature.     

A clustering algorithm for supplier base management

Supplier base management is an important strategy for managing global, customer driven supply chains. Successful supplier base management can effectively handle supply side exceptions, which may have significant business implications. Currently, there is a trend to reduce the size of the supplier base which makes the coordination and interaction among suppliers more effective, less costly and time consuming. The goal of this research is to present a clustering algorithm, named min-min-roughness (MMR) to cluster suppliers into smaller, more manageable groups with similar characteristics. Due to the fact that supplier data are mainly categorical in nature, MMR, based on rough set theory (RST) is developed for categorical data clustering which is also capable of handling the uncertainty during the clustering process. One potential benefit of applying MMR to supplier base management is that more realistic benchmarking can be obtained and the fulfilment operation can be sped up by reducing the number of variables impacting the operations. In addition, the characteristics of each smaller group of suppliers can be summarised and exploited to handle supply side disruptions.     

一种快速移动对象轨道聚类算法

针对已有轨道聚类(TRACLUS)算法的线段聚类模块需要对划分后的每条线段进行邻域查询的问题,将取样技术引入轨道聚类,提出一种快速移动对象轨道聚类(FTCS)算法。FTCS算法根据基于极大连通子图的合并原理,对核心线段的Eps邻域以及与该Eps邻域相重叠的所有轨道聚类进行合并,避免了TRACLUS算法中核心线段Eps邻域内线段的不必要邻域查询操作。在真实和合成轨道数据集上的大量实验结果表明,FTCS算法显著降低了邻域查询操作次数,在保持TRACLUS算法轨道聚类质量的同时,成倍提高了轨道聚类的时间效率。     

一种基于GA的聚类集成算法

提出了一种基于GA的聚类集成算法ECUNGA(ensemble clustering using NMI and GA).算法利用GA搜索一个与聚类集体差异度小的聚类,以此来达到综合聚类集体信息,得到更优秀的聚类的目的.算法相比于传统基于互信息理论的方法,使用GA搜索,提高了搜索的能力且具有较低计算复杂度.最后,在UCI数据集上进行实验,取得了理想的效果.     

Parallel architecture for cotton crop classification using WLI-Fuzzy clustering algorithm and Bs-Lion neural network model

The application of remote sensory images in crop monitoring has been increasing in the recent years due to its high classification accuracy. In this paper, a novel parallel classification methodology is proposed using a new clustering and classification concept. A novel neural network model with the Bs-Lion training algorithm is developed by integrating the Bayesian regularization training with the Lion Algorithm. Here, two levels of parallel processing are performed, namely parallel WLI-Fuzzy clustering and parallel BS-Lion neural network classification. The experimentation of the proposed parallel methodology is carried out using satellite images obtained from the Indian remote sensing satellite IRS-P6. The performance of the proposed system is compared with the existing techniques using validation measures accuracy, sensitivity and specificity. The experimentations resulted in promising results with an accuracy of 0.8994, sensitivity of 0.8682 and specificity of 0.8739, which favour the performance of the proposed parallel architecture in the classification.