Fast self-organizing feature map algorithm

We present an efficient approach to forming feature maps. The method involves three stages. In the first stage, we use the K-means algorithm to select N(2) (i.e., the size of the feature map to be formed) cluster centers from a data set. Then a heuristic assignment strategy is employed to organize the N(2) selected data points into an NxN neural array so as to form an initial feature map. If the initial map is not good enough, then it will be fine-tuned by the traditional Kohonen self-organizing feature map (SOM) algorithm under a fast cooling regime in the third stage. By our three-stage method, a topologically ordered feature map would be formed very quickly instead of requiring a huge amount of iterations to fine-tune the weights toward the density distribution of the data points, which usually happened in the conventional SOM algorithm. Three data sets are utilized to illustrate the proposed method.     

自组织神经网络的参数自适应方法

自组织神经网络的主要目的是将任意维数的输入信号模式转变成为一维或二维的离散映射,并且以拓扑有序的方式自适应地实现这个过程。学习过程中,对邻域宽度函数和学习率函数参数是根据经验选择的,没有一定的规则或方法,因此,邻域保持映射的获得往往先于参数的学习过程。将线性Kalman滤波器和基于无先导变换的Kalman滤波器分别用于学习率函数和邻域宽度函数的预测,可以提高自组织神经网络的学习能力。改进后的算法可以根据输入数据自适应地调整邻域宽度函数和学习率函数。     

Constrained self-organizing feature map to preserve feature extraction topology

In many classification problems, it is necessary to consider the specific location of an n-dimensional space from which features have been calculated. For example, considering the location of features extracted from specific areas of a two-dimensional space, as an image, could improve the understanding of a scene for a video surveillance system. In the same way, the same features extracted from different locations could mean different actions for a 3D HCI system. In this paper, we present a self-organizing feature map able to preserve the topology of locations of an n-dimensional space in which the vector of features have been extracted. The main contribution is to implicitly preserving the topology of the original space because considering the locations of the extracted features and their topology could ease the solution to certain problems. Specifically, the paper proposes the n-dimensional constrained self-organizing map preserving the input topology (nD-SOM-PINT). Features in adjacent areas of the n-dimensional space, used to extract the feature vectors, are explicitly in adjacent areas of the nD-SOM-PINT constraining the neural network structure and learning. As a study case, the neural network has been instantiate to represent and classify features as trajectories extracted from a sequence of images into a high level of semantic understanding. Experiments have been thoroughly carried out using the CAVIAR datasets (Corridor, Frontal and Inria) taken into account the global behaviour of an individual in order to validate the ability to preserve the topology of the two-dimensional space to obtain high-performance classification for trajectory classification in contrast of non-considering the location of features. Moreover, a brief example has been included to focus on validate the nD-SOM-PINT proposal in other domain than the individual trajectory. Results confirm the high accuracy of the nD-SOM-PINT outperforming previous methods aimed to classify the same datasets.

     

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.     

基于自组织特征映射的聚类集成算法

为改善单一聚类算法的聚类性能,提出一种基于自组织特征映射(SOM)的聚类集成算法.该算法利用多个具有差异性的聚类成员,将原始数据集转换成一个新的特征空间矩阵;然后计算各个聚类成员的聚类综合质量,并将其作为新特征空间矩阵的属性权重,最后利用SOM神经网络进行集成,产生最终的共识聚类结果.实验结果表明,与集成前的基聚类算法和其它聚类集成算法相比,该算法能够有效地提高聚类质量.     

Integration of self-organizing feature map and K-means algorithm for market segmentation

Cluster analysis is a common tool for market segmentation. Conventional research usually employs the multivariate analysis procedures. In recent years, due to their high performance in engineering, artificial neural networks have also been applied in the area of management. Thus, this study aims to compare three clustering methods: (1) the conventional two-stage method, (2) the self-organizing feature maps and (3) our proposed two-stage method, via both simulated and real-world data. The proposed two-stage method is a combination of the self-organizing feature maps and the K-means method. The simulation results indicate that the proposed scheme is slightly better than the conventional two-stage method with respect to the rate of misclassification, and the real-world data on the basis of Wilk's Lambda and discriminant analysis.Scope and purposeThe general idea of segmentation, or clustering, is to group items that are similar. A commonly used method is the multivariate analysis [4]. These methods consist of hierarchical methods, like Ward's minimum variance method, and the non-hierarchical methods, such as the K-means method. Owing to increase in computer power and decrease in computer costs, artificial neural networks (ANNs), which are distributed and parallel information processing systems successfully applied in the area of engineering, have recently been employed to solve the marketing problems. This study aims to discuss the possibility of integrating ANN and multivariate analysis. A two-stage method, which first uses the self-organizing feature maps to determine the number of clusters and the starting point and then employs the K-means method to find the final solution, is proposed. This method provides the marketing analysts a more sophisticated way to analyze the consumer behavior and determine the marking strategy. A case study is also employed to demonstrate the validity of the proposed method.     

Extraction of rice-planted area using a self-organizing feature map

We introduce a neural network of self-organizing feature map (SOM) to classify remote-sensing data, including microwave and optical sensors, for the estimation of areas of planted rice. This method is an unsupervised neural network which has the capability of nonlinear discrimination, and the classification function is determined by learning. The satellite data are observed before and after rice planting in 1999. Three sets of RADARSAT and one set of SPOT/HRV data were used in Higashi–Hiroshima, Japan. The RADARSAT image has only one band of data and it is difficult to extract the rice-planted area. However, the SAR back-scattering intensity in a rice-planted area decreases from April to May and increases from May to June. Therefore, three RADARSAT images from April to June were used in this study. The SOM classification was applied the RADARSAT and SPOT data to evaluate the rice-planted area estimation. It is shown that the SOM is useful for the classification of satellite data.     

Edge detection in multispectral images using the self-organizing map

In this paper, two new methods for edge detection in multispectral images are presented. They are based on the use of the self-organizing map (SOM) and a grayscale edge detector. With the 2-dimensional SOM the ordering of pixel vectors is obtained by applying the Peano scan, whereas this can be omitted using the 1-dimensional SOM. It is shown that using the R-ordering based methods some parts of the edges may be missed. However, they can be found using the proposed methods. Using them it is also possible to find edges in images which consist of metameric colors. Finally, it is shown that the proposed methods find the edges properly from real multispectral airplane images. The size of the SOM determines the amount of found edges. If the SOM is taught using a large color vector database, the same SOM can be utilized for numerous images.     

一种基于彩色图像的道路交通标志检测新方法

提出了一种改进的彩色图像分割方法,并将该方法与不变矩理论相结合用于检测彩色图像中的交通禁令标志。首先对采集图像进行预处理,包括对图像进行RGB到HSI或改进HSI颜色空间的转换和图像形态学的运算,然后对图像中不同的封闭子区域进行标记,并去除不满足面积阈值的子区域。分别计算剩下子区域的hu矩组得到每个子区域的７个图像特征值。将相应子区域的特征值与事先准备好的环形和三角形路标特征值用欧式距离分类器进行比较判别。实验结果表明,此方法能准确并较为快速地实现警告标志检测。     

Semi-supervised change detection using modified self-organizing feature map neural network

In the present article, semi-supervised learning is integrated with an unsupervised context-sensitive change detection technique based on modified self-organizing feature map (MSOFM) network. In the proposed methodology, training of the MSOFM network is initially performed using only a few labeled patterns. Thereafter, the membership values, in both the classes, for each unlabeled pattern are determined using the concept of fuzzy set theory. The soft class label for each of the unlabeled patterns is then estimated using the membership values of its K nearest neighbors. Here, training of the network using the unlabeled patterns along with a few labeled patterns is carried out iteratively. A heuristic method has been suggested to select some patterns from the unlabeled ones for training. To check the effectiveness of the proposed methodology, experiments are conducted on three multi-temporal and multi-spectral data sets. Performance of the proposed work is compared with that of two unsupervised techniques, a supervised technique and two semi-supervised techniques. Results are also statistically validated using paired t-test. The proposed method produced promising results.     

Generalizing self-organizing map for categorical data

The self-organizing map (SOM) is an unsupervised neural network which projects high-dimensional data onto a low-dimensional grid and visually reveals the topological order of the original data. Self-organizing maps have been successfully applied to many fields, including engineering and business domains. However, the conventional SOM training algorithm handles only numeric data. Categorical data are usually converted to a set of binary data before training of an SOM takes place. If a simple transformation scheme is adopted, the similarity information embedded between categorical values may be lost. Consequently, the trained SOM is unable to reflect the correct topological order. This paper proposes a generalized self-organizing map model that offers an intuitive method of specifying the similarity between categorical values via distance hierarchies and, hence, enables the direct process of categorical values during training. In fact, distance hierarchy unifies the distance computation of both numeric and categorical values. The unification is done by mapping the values to distance hierarchies and then measuring the distance in the hierarchies. Experiments on synthetic and real datasets were conducted, and the results demonstrated the effectiveness of the generalized SOM model.     

Conformal self-organizing map for a genus-zero manifold

This paper presents the implementation of a surface mesh on a genus-zero manifold with 3D scattered data of sculpture surfaces using the conformal self-organizing map (CSM). It starts with a regular mesh on a sphere and gradually shapes the regular mesh to match its object’s surface by using the CSM. It can drape a uniform mesh on an object with a high degree of conformality. It accomplishes the surface reconstruction and also defines a conformal mapping from a sphere to the object’s manifold.     

Investigation of self-organizing map for genetic algorithm

This paper describes self-organizing maps for genetic algorithm (SOM-GA) which is the combinational algorithm of a real-coded genetic algorithm (RCGA) and self-organizing map (SOM). The self-organizing maps are trained with the information of the individuals in the population. Sub-populations are defined by the help of the trained map. The RCGA is performed in the sub-populations. The use of the sub-population search algorithm improves the local search performance of the RCGA. The search performance is compared with the real-coded genetic algorithm (RCGA) in three test functions. The results show that SOM-GA can find better solutions in shorter CPU time than RCGA. Although the computational cost for training SOM is expensive, the results show that the convergence speed of SOM-GA is accelerated according to the development of SOM training.     

The parameterless self-organizing map algorithm

The parameterless self-organizing map (PLSOM) is a new neural network algorithm based on the self-organizing map (SOM). It eliminates the need for a learning rate and annealing schemes for learning rate and neighborhood size. We discuss the relative performance of the PLSOM and the SOM and demonstrate some tasks in which the SOM fails but the PLSOM performs satisfactory. Finally we discuss some example applications of the PLSOM and present a proof of ordering under certain limited conditions.     

Clustering of the self-organizing map

The self-organizing map (SOM) is an excellent tool in exploratory phase of data mining. It projects input space on prototypes of a low-dimensional regular grid that can be effectively utilized to visualize and explore properties of the data. When the number of SOM units is large, to facilitate quantitative analysis of the map and the data, similar units need to be grouped, i.e., clustered. In this paper, different approaches to clustering of the SOM are considered. In particular, the use of hierarchical agglomerative clustering and partitive clustering using K-means are investigated. The two-stage procedure-first using SOM to produce the prototypes that are then clustered in the second stage-is found to perform well when compared with direct clustering of the data and to reduce the computation time.     

Some novel classifiers designed using prototypes extracted by a newscheme based on self-organizing feature map

We propose two new comprehensive schemes for designing prototype-based classifiers. The scheme addresses all major issues (number of prototypes, generation of prototypes, and utilization of the prototypes) involved in the design of a prototype-based classifier. First we use Kohonen's self-organizing feature map (SOFM) algorithm to produce a minimum number (equal to the number of classes) of initial prototypes. Then we use a dynamic prototype generation and tuning algorithm (DYNAGEN) involving merging, splitting, deleting, and retraining of the prototypes to generate an adequate number of useful prototypes. These prototypes are used to design a "1 nearest multiple prototype (1-NMP)" classifier. Though the classifier performs quite well, it cannot reasonably deal with large variation of variance among the data from different classes. To overcome this deficiency we design a "1 most similar prototype (1-MSP)" classifier. We use the prototypes generated by the SOFM-based DYNAGEN algorithm and associate with each of them a zone of influence. A norm (Euclidean)-induced similarity measure is used for this. The prototypes and their zones of influence are fine-tuned by minimizing an error function. Both classifiers are trained and tested using several data sets, and a consistent improvement in performance of the latter over the former has been observed. We also compared our classifiers with some benchmark results available in the literature.     

A self-organizing feature map for clustering nucleic acids Application to a data matrix containing A-DNA and B-DNA dinucleotides

A self-organizing feature map to cluster DNA dinucleotides is presented. During a training session 244 training patterns, each consisting of nine torsion angles, are clustered in a 10 by 10 map. The method is successful for separating the four known DNA classes in the training set. Contour plots of the weights after a training session indicate gradients in torsion angles corresponding to class separation. Moreover, certain units in the map probably correspond to unfavourable torsion angle combinations resulting in, e.g. van der Waals clashes. Hence, although no direct relation to a conformation's energy (as in a Ramachandran plot) is present in the map, it may provide a multidimensional interpretation of accessible and forbidden areas for dinucleotides. The applicability of the method on this DNA data matrix shows its potential to be used in more extensive structural analysis studies, e.g. in a case of comparing DNA with RNA. Several test patterns resulting from molecules with unusual structural characteristics are identified with the map.     

The temporal correlation hypothesis for self-organizing feature maps

Feature maps, in which one or more aspects of the environment are systematically represented over the surface of the cerebral cortex, are often found in primary sensory and motor cortical regions of the vertebrate brain. They have inspired a great deal of computational modelling, and this has provided evidence that such maps are emergent properties of the interactions of numerous cortical neurons and their adaptive, nonlinear connections. In this paper, we address the issue of how multiple feature maps that coexist in the same region of cerebral cortex align with each other. We hypothesize that such alignment is governed by temporal correlations: features in one map that are temporally correlated with those in another come to occupy the same spatial locations over time. To examine the feasibility of this hypothesis and to establish some of its detailed implications, we initially studied a computational model of primary sensorimotor cortex. Coexisting sensory and motor maps formed and generally aligned in a fashion consistent with the temporal correlation hypothesis. We summarize these results, and then mathematically analyse a simplified model of self-organization during unsupervised learning. We show that the properties observed computationally are quite general: that temporally correlated inputs become spatially correlated (i.e. aligned), while input patterns that are temporally anti-correlated tend to result in mutually exclusive (i.e. unaligned) spatial distributions. This work provides a framework in which to interpret and understand future experimental studies of map relationships.