自组织特征映射网络在故障诊断中的应用

运用自组织特征映射神经网络的工作原理和具体实现算法进行故障诊断分析,在对已有神经网络聚类分析方法概括和总结的基础上,结合实验数据、仿真数据对自组织特征映射算法故障模型诊断进行研究,得出了有意义的结论.     

Effects of varying parameters on properties of self-organizing feature maps

The behavior of self-organizing feature maps is critically dependent on parameters such as lateral connection radius, lateral inhibition intensity, and network size. With no theoretical guidelines for the choice of these parameters, they are usually selected through a trial-and-error process. In order to provide heuristic guidelines for future model designers as well as to give insight into which model features are responsible for specific aspects of maps, we systematically varied these parameters and studied their effects on the properties of a self-organizing feature map. The connectivity radius was found to determine the size of activation clusters quadratically. As the intensity of lateral inhibition was varied, feature patterns varied from stripe-like to clusters in the map, with other intermediate patterns also occurring. The number of clusters of each feature increased nonlinearly as the network size increased.     

Qualitative analysis of goat and sheep production data using self-organizing maps

Abstract: The aim of this study was to analyse the relationship between different small ruminant livestock production systems with different levels of specialization. The analysis is carried out by using the self-organizing map. This tool allows high-dimensional input spaces to be mapped into much lower-dimensional spaces, thus making it much more straightforward to understand any set of data. These representations enable the visual extraction of qualitative relationships among variables (visual data mining), converting the data to maps. The data used in this study were obtained from surveys completed by farmers who are principally dedicated to goat and sheep production. With the self-organizing map we found a relationship between qualitative and quantitative variables showing that more specialized farms have greater milk incomes per goat, highlighting farms that have a greater number of animals, better facilities (including milking machines) or animals fed with elaborated diets. The use of self-organizing maps for the analysis of this kind of data has proven to be highly valuable in extracting qualitative conclusions and in guiding improvements in farm performance.     

I-Miner环境下聚类及可视化研究

聚类分析是数据挖掘中的核心技术,利用相关的可视化方法显示聚类结果,将数据分布以直观、形象的图形方式呈现给决策者,使得决策者可以直观地分析数据。I-Miner是一个企业级的数据挖掘工具,利用I-Miner软件进行聚类分析,并用多种方法将聚类结果可视化。通过S语言拓展软件功能,编程实现了K-Medoid算法、SOM算法、SOM与K-Medoids结合的聚类组合算法,尤其是在高维数据的可视化上,实现了星图法和SOM之U矩阵法,弥补软件中聚类和可视化模块较少的不足。     

Spatial data mining and visualization based on self-organizing map

Based on the traditional spatial data analysis, a novel mode of spatial data mining and visualization is proposed which integrates the self-organizing map for the actual problem. Simulations for IRIS data show that this method （computational and visual） can collaboratively discover complex pattems in large spatial datasets, in an effective and efficient way.     

改进的基于SOM的高维数据可视化算法

通过对高维数据可视化方法的系统研究,提出了一种新的基于自组织映射（Self-Organizing Map,SOM）的算法。为了表现该方法的特点,将其称为三维自组织映射（Three-Dimensional SOM,TDSOM）。它在对高维数据记录集进行SOM分析后将其投影到三维坐标系中的特定的点集上,最终形成三维模型。该模型弥补了传统模型难以清晰准确地展现高维数据的缺陷,并且新模型着重于在一个比二维平面更为广阔的三维立体空间中展现海量数据。使用者通常可以根据当前领域的专业知识在分析模型的基础上得出有意义的模式。新方法可以广泛使用在数据挖掘和模式识别等领域。     

A self-organizing feature maps and data mining based decision support system for liability authentications of traffic crashes

This study develops a decision support tool for liability authentications of two-vehicle crashes based on generated self-organizing feature maps (SOM) and data mining (DM) models. Factors critical to liability attributions commonly identified theoretically and practically were first selected. Both SOM and DM models were then generated for frontal, side, and rear collisions of two-vehicle crashes. Appropriateness of all generated models was evaluated and confirmed. Finally, a decision support tool was developed using active server pages. Although with small data size, the decision support system was considered capable of giving reasonably good liability attributions and references on given cases.     

Research on the clustering algorithm of ocean big data based on self-organizing neural network

In the construction of a smart marine, marine big data mining has a significant impact on the growing maritime industry in the Beibu Gulf. Clustering is the key technology of marine big data mining, but the conventional clustering algorithm cannot achieve the efficient clustering of marine data. According to the characteristics of marine big data, a marine big data clustering scheme based on self-organizing neural network (SOM) algorithm is proposed. First, the working principle of SOM algorithm is analyzed, and the algorithm's two-dimensional network model, similarity model and competitive learning model are focused. Secondly, combining with the working principle of algorithm, the marine big data clustering process and algorithm achievement based on SOM algorithm are developed; finally, experiments show that all vectors in marine big data clustering are stable, and the neurons in the output layer of clustering result have obvious consistency with the data itself, which shows the effectiveness of SOM algorithm in marine big data clustering.     

Process control of time-varying systems using parameter-less self-organizing maps

Traditional control charts, such as Hotelling’s T², are effective in detecting abnormal patterns. However, most control charts do not take into account a time-varying property in a process. In the present study, we propose a parameter-less self-organizing map-based control chart that can handle a situation in which changes occur in the distribution or parameter of the target observations. The control limits of the proposed chart are determined by estimating the empirical level of significance on the percentile using the bootstrap method. Experimental results obtained by using simulated data and actual process data from the manufacturing process for a thin-film transistor-liquid crystal display demonstrate the effectiveness and usefulness of the proposed algorithm.     

Categorical data visualization and clustering using subjective factors

Clustering is an important data mining problem. However, most earlier work on clustering focused on numeric attributes which have a natural ordering to their attribute values. Recently, clustering data with categorical attributes, whose attribute values do not have a natural ordering, has received more attention. A common issue in cluster analysis is that there is no single correct answer to the number of clusters, since cluster analysis involves human subjective judgement. Interactive visualization is one of the methods where users can decide a proper clustering parameters. In this paper, a new clustering approach called CDCS (Categorical Data Clustering with Subjective factors) is introduced, where a visualization tool for clustered categorical data is developed such that the result of adjusting parameters is instantly reflected. The experiment shows that CDCS generates high quality clusters compared to other typical algorithms.     

GPROC—an integrated system for the processing of numerical scientific data

Work done on an interactive graphical processing system is described. An overview is given of the uses to which a system could be put, with especial reference to aspects of solid state physics. Algorithms which may be usefully included in such a system, when used in a technical scientific environment, are described. The problems of implementing such a system on small microcomputer systems are outlined, and some possible areas of future development are discussed.     

Visualization and interaction of finite element analysis in augmented reality

In this paper, augmented reality (AR) is used to enhance the visualization and interaction of finite element analysis (FEA) of structures. An integrated simulation system is proposed which acquires input data using sensors and uses AR technology to visualize FEA results in the real world. A number of intuitive interaction methods have been devised and implemented in this system. The user can perform real-time FEA simulation to investigate structural behavior under different loading conditions either through manipulating virtual loads or creating different loading conditions. Exploration of FEA results is enhanced through natural interfaces for manipulating, slicing and clipping the result data. Moreover, the user can modify the FE models of the structures through simplified operations for different purposes, e.g., adding structural members for stiffening and performing local mesh refinement. The modified model can be re-analyzed automatically. A prototype system has been built and a case study has been implemented to demonstrate the innovative interaction methods and evaluate the system performance.     

Extending parallelization of the self-organizing map by combining data and network partitioned methods

High-dimensional data is pervasive in many fields such as engineering, geospatial, and medical. It is a constant challenge to build tools that help people in these fields understand the underlying complexities of their data. Many techniques perform dimensionality reduction or other “compression” to show views of data in either two or three dimensions, leaving the data analyst to infer relationships with remaining independent and dependent variables. Contextual self-organizing maps offer a way to represent and interact with all dimensions of a data set simultaneously. However, computational times needed to generate these representations limit their feasibility to realistic industry settings. Batch self-organizing maps provide a data-independent method that allows the training process to be parallelized and therefore sped up, saving time and money involved in processing data prior to analysis. This research parallelizes the batch self-organizing map by combining network partitioning and data partitioning methods with CUDA on the graphical processing unit to achieve significant training time reductions. Reductions in training times of up to twenty-five times were found while using map sizes where other implementations have shown weakness. The reduced training times open up the contextual self-organizing map as viable option for engineering data visualization.     

Visual exploration of production data using small multiples design with non-uniform color mapping

Visual data mining may overcome some of the flexibility problem often suffered by computer-centered data mining approaches. This can happen because human beings are introduced to the information discovery loop to take advantage of their natural strength in creative thinking and rapid visual pattern recognition to discover information not defined a priori and to perform approximated reasoning that computer algorithms are hard to do. This paper presents a novel visual exploration approach for mining abstract, multi-dimensional data stored in tables in a relational database. The visual image is constructed by converting each table into a visualization unit called a table graph and then assembling these table graphs together to form a small multiples design. Different types of non-uniform color mappings to render this small multiples design could be automatically generated by minimizing the weight differences of colors in the visual image. These non-uniform color mappings are designed in such a way that the adjacent glyphs in a table graph that have near underlying values will be assigned with the same color. As such, visual patterns not able to see under the traditional uniform color mapping could be revealed. This enables the users to examine the input tables from different perspectives. The proposed flexible visualization method has been applied to generate visual images from which the users could quickly and easily compare the machine idle cost performances of alternative master production plans.     

基于SOM和引力场聚类的金融数据可视化

平行坐标技术是信息可视化中重要的分析手段,可以实现多维数据在二维空间上的可视化.为了给用户提供一种快捷、方便的金融数据可视化及分析工具,提出一种基于引力场聚类的金融数据可视化方法.首先利用自组织映射(SOM)对初始金融数据进行分类,使每类数据都含有特定的经济意义;然后进行视觉聚类,利用引力场原理对每个类中的折线进行聚拢,对类与类之间进行排斥,再通过设置不透明度以及交互操作等手段对可视化结果进行增强.实验结果表明,该方法可以形成清晰的可视化聚类结果,便于发现数据的变化规律.     

图像纹理特征数据挖掘的不确定性分析

不确定性分析是数据挖掘与知识发现的重要内容,对图像纹理特征数据挖掘的基本原理进行了分析、解释,从问题的求解、数据的产生、挖掘的过程以及最终的结果等几个角度分析了图像纹理特征数据挖掘中的不确定性,并讨论了目前不确定性问题分析与处理的若干方法,引入不确定性分析的有力工具--云模型,研究图像纹理特征数据挖掘中的不确定性,并给出了实验结果及分析.     

Handling of incomplete data sets using ICA and SOM in data mining

Based on independent component analysis (ICA) and self-organizing maps (SOM), this paper proposes an ISOM-DH model for the incomplete data’s handling in data mining. Under these circumstances the data remain dependent and non-Gaussian, this model can make full use of the information of the given data to estimate the missing data and can visualize the handled high-dimensional data. Compared with mixture of principal component analyzers (MPCA), mean method and standard SOM-based fuzzy map model, ISOM-DH model can be applied to more cases, thus performing its superiority. Meanwhile, the correctness and reasonableness of ISOM-DH model is also validated by the experiment carried out in this paper.     

MRCCA: A novel CCA based method and its application in feature extraction and fusion for matrix data

Multiset features extracted from the same pattern usually represent different characteristics of data, meanwhile, matrices or 2-order tensors are common forms of data in real applications. Hence, how to extract multiset features from matrix data is an important research topic for pattern recognition. In this paper, by analyzing the relationship between CCA and 2D-CCA, a novel feature extraction method called multiple rank canonical correlation analysis (MRCCA) is proposed, which is an extension of 2D-CCA. Different from CCA and 2D-CCA, in MRCCA k pairs left transforms and k pairs right transforms are sought to maximize correlation. Besides, the multiset version of MRCCA termed as multiple rank multiset canonical correlation analysis (MRMCCA) is also developed. Experimental results on five real-world data sets demonstrate the viability of the formulation, they also show that the recognition rate of our method is higher than other methods and the computing time is competitive.     

Hyperspectral discrimination of tropical dry forest lianas and trees: Comparative data reduction approaches at the leaf and canopy levels

A dataset of spectral signatures (leaf level) of tropical dry forest trees and lianas and an airborne hyperspectral image (crown level) are used to test three hyperspectral data reduction techniques (principal component analysis, forward feature selection and wavelet energy feature vectors) along with pattern recognition classifiers to discriminate between the spectral signatures of lianas and trees. It was found at the leaf level the forward waveband selection method had the best results followed by the wavelet energy feature vector and a form of principal component analysis. For the same dataset our results indicate that none of the pattern recognition classifiers performed the best across all reduction techniques, and also that none of the parametric classifiers had the overall lowest training and testing errors. At the crown level, in addition to higher testing error rates (7%), it was found that there was no optimal data reduction technique. The significant wavebands were also found to be different between the leaf and crown levels. At the leaf level, the visible region of the spectrum was the most important for discriminating between lianas and trees whereas at the crown level the shortwave infrared was also important in addition to the visible and near infrared.