The Self Organizing Map as a Tool for Cluster Analysis

2003

Self-Organizing Map (SOM) is a special kind of unsupervised neural network. SOM consists of regular, usually two-dimensional, neurons. During training, SOM forms an elastic net that folds onto the "cloud" formed by the input data. Thus, SOM can be interpreted as a topology preserving mapping from input space onto the twodimensional grid of neurons. In mining data, SOM has been used as a clustering technique. As there are several important issues concerning with data mining clustering techniques, some experiments have been done with the goal of discovering the relation between SOM and the issues. This paper discusses SOM, the experiments and the analytical results of how SOM, in some way, has provided good solutions to several of the issues.

Abstract—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 -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.

Data mining is generally the process of examining data from different aspects and summarizing it into valuable information. There are number of data mining software's for analysing the data. They allow users to examine the data from various angles, categorize it, and summarize the relationships identified.

Lecture Notes in Computer Science, 2005

This paper presents an innovative, adaptive variant of Kohonen's selforganizing maps called ASOM, which is an unsupervised clustering method that adaptively decides on the best architecture for the self-organizing map. Like the traditional SOMs, this clustering technique also provides useful information about the relationship between the resulting clusters. Applications of the resulting software to clustering biological data are discussed in detail.

Information Sciences, 2004

The Self-Organizing Map (SOM) is a powerful tool in the exploratory phase of data mining. It is capable of projecting high-dimensional data onto a regular, usually 2-dimensional grid of neurons with good neighborhood preservation between two spaces. However, due to the dimensional conflict, the neighborhood preservation cannot always lead to perfect topology preservation. In this paper, we establish an Expanding SOM (ESOM) to preserve better topology between the two spaces. Besides the neighborhood relationship, our ESOM can detect and preserve an ordering relationship using an expanding mechanism. The computation complexity of the ESOM is comparable with that of the SOM. Our experiment results demonstrate that the ESOM constructs better mappings than the classic SOM, especially, in terms of the topological error. Furthermore, clustering results generated by the ESOM are more accurate than those obtained by the SOM.

2005

One of the most widely used clustering techniques used in GISc problems is the k-means algorithm. One of the most important issues in the correct use of k-means is the initialization procedure that ultimately determines which part of the solution space will be searched. In this paper we briefly review different initialization procedures, and propose Kohonen's Self-Organizing Maps as the most convenient method, given the proper training parameters.

Computational Statistics & Data Analysis, 2001

The self-organizing map (SOM) network was originally designed for solving problems that involve tasks such as clustering, visualization, and abstraction. While Kohonen's SOM networks have been successfully applied as a classiÿcation tool to various problem domains, their potential as a robust substitute for clustering and visualization analysis remains relatively unresearched. We believe the inadequacy of attention in the research and application of using SOM networks as a clustering method is due to its lack of procedures to generate groupings from the SOM output. In this paper, we extend the original Kohonen SOM network to include a contiguity-constrained clustering method to perform clustering based on the output map generated by the network. We compare the result with that of the other clustering tools using a classic problem from the domain of group technology. The result shows that the combination of SOM and the contiguity-constrained clustering method produce clustering results that are comparable with that of the other clustering methods. We further test the applicability of the method with two widely referenced machine-learning cases and compare the results with that of several popular statistical clustering methods.

2012

As a special class of artificial neural networks the Self Organizing Map is used extensively as a clustering and visualization technique in exploratory data analysis. This chapter provides a general introduction to the structure, algorithm and quality of Self Organizing Maps and presents industrial engineering related applications reported in the literature.

Self-Organizing Map (SOM) is a neural network model which is used to obtain a topology-preserving mapping from the (usually high dimensional) input/feature space to an output/map space of fewer dimensions (usually two or three in order to facilitate visualization). Neurons in the output space are connected with each other but this structure remains fixed throughout training and learning is achieved through the updating of neuron reference vectors in feature space. Despite the fact that growing variants of SOM overcome the fixed structure limitation they increase computational cost and also do not allow the removal of a neuron after its introduction. In this paper, a variant of SOM is proposed called AMSOM (Adaptive Moving Self-Organizing Map) that on the one hand creates a more flexible structure where neuron positions are dynamically altered during training and on the other hand tackles the drawback of having a predefined grid by allowing neuron addition and/or removal during training. Experiments using multiple literature datasets show that the proposed method improves training performance of SOM, leads to a better visualization of the input dataset and provides a framework for determining the optimal number and structure of neurons.

2005

In this paper, we propose a new clustering method consisting in automated “flood- fill segmentation” of the U*-matrix of a Self-Organizing Map after training. Using several artificial datasets as a benchmark, we find that the clustering results of our U*F method are good over a wide range of critical dataset types. Furthermore, comparison to standard clustering algorithms (K-means, single-linkage and Ward) directly applied on the same datasets show that each of the latter performs very bad on at least one kind of dataset, contrary to our U*F clustering method: while not always the best, U*F clustering has the great advantage of exhibiting consistently good results. Another advantage of U*F is that the computation cost of the SOM segmentation phase is negligible, contrary to other SOM-based clustering approaches which apply O(n2logn) standard clustering algorithms to the SOM prototypes. Finally, it should be emphasized that U*F clustering does not require a priori knowledge on the nu...