Self Organized Map

In this paper, we introduce a general modeling technique, called vector-quantized temporal associative memory (VQTAM), which uses Kohonen's self-organizing map (SOM) as an alternative to multilayer perceptron (MLP) and radial basis function (RBF) neural models for dynamical system identification and control. We demonstrate that the estimation errors decrease as the SOM training proceeds, allowing the VQTAM scheme to be understood as a self-supervised gradient-based error reduction method. The performance of the proposed approach is evaluated on a variety of complex tasks, namely: i) time series prediction; ii) identification of SISO/MIMO systems; and iii) nonlinear predictive control. For all tasks, the simulation results produced by the SOM are as accurate as those produced by the MLP network, and better than those produced by the RBF network. The SOM has also shown to be less sensitive to weight initialization than MLP networks. We conclude the paper by discussing the main properties of the VQTAM and their relationships to other well established methods for dynamical system identification. We also suggest directions for further work.

Array technologies have made it straightforward to monitor simultaneously the expression pattern of thousands of genes. The challenge now is to interpret such massive data sets. The first step is to extract the fundamental patterns of gene expression inherent in the data. This paper describes the application of self-organizing maps, a type of mathematical cluster analysis that is particularly well suited for recognizing and classifying features in complex, multidimensional data. The method has been implemented in a publicly available computer package, GENECLUSTER, that performs the analytical calculations and provides easy data visualization. To illustrate the value of such analysis, the approach is applied to hematopoietic differentiation in four well studied models (HL-60, U937, Jurkat, and NB4 cells). Expression patterns of some 6,000 human genes were assayed, and an online database was created. GENECLUSTER was used to organize the genes into biologically relevant clusters that suggest novel hypotheses about hematopoietic differentiation-for example, highlighting certain genes and pathways involved in ''differentiation therapy'' used in the treatment of acute promyelocytic leukemia.

We describe and demonstrate an algorithm that takes as input an unorganized set of points fx1; : : : ; x n g IR 3 on or near an unknown manifold M, and produces as output a simplicial surface that approximates M. Neither the topology, the presence of boundaries, nor the geometry of M are assumed to be known in advance -all are inferred automatically from the data. This problem naturally arises in a variety of practical situations such as range scanning an object from multiple view points, recovery of biological shapes from two-dimensional slices, and interactive surface sketching.

Faces represent complex, multidimensional, meaningful visual stimuli and developing a computational model for face recognition is difficult [43]. We present a hybrid neural network solution which compares favorably with other methods. The system combines local image sampling, a self-organizing map neural network, and a convolutional neural network. The self-organizing map provides a quantization of the image samples into a topological space where inputs that are nearby in the original space are also nearby in the output space, thereby providing dimensionality reduction and invariance to minor changes in the image sample, and the convolutional neural network provides for partial invariance to translation, rotation, scale, and deformation. The convolutional network extracts successively larger features in a hierarchical set of layers. We present results using the Karhunen-Loève transform in place of the self-organizing map, and a multi-layer perceptron in place of the convolutional network. The Karhunen-Loève transform performs almost as well (5.3% error versus 3.8%). The multi-layer perceptron performs very poorly (40% error versus 3.8%). The method is capable of rapid classification, requires only fast, approximate normalization and preprocessing, and consistently exhibits better classification performance than the eigenfaces approach [43] on the database considered as the number of images per person in the training database is varied from 1 to 5. With 5 images per person the proposed method and eigenfaces result in 3.8% and 10.5% error respectively. The recognizer provides a measure of confidence in its output and classification error approaches zero when rejecting as few as 10% of the examples. We use a database of 400 images of 40 individuals which contains quite a high degree of variability in expression, pose, and facial details. We analyze computational complexity and discuss how new classes could be added to the trained recognizer.

We introduce a method of functionally classifying genes by using gene expression data from DNA microarray hybridization experiments. The method is based on the theory of support vector machines (SVMs). SVMs are considered a supervised computer learning method because they exploit prior knowledge of gene function to identify unknown genes of similar function from expression data. SVMs avoid several problems associated with unsupervised clustering methods, such as hierarchical clustering and self-organizing maps. SVMs have many mathematical features that make them attractive for gene expression analysis, including their flexibility in choosing a similarity function, sparseness of solution when dealing with large data sets, the ability to handle large feature spaces, and the ability to identify outliers. We test several SVMs that use different similarity metrics, as well as some other supervised learning methods, and find that the SVMs best identify sets of genes with a common function using expression data. Finally, we use SVMs to predict functional roles for uncharacterized yeast ORFs based on their expression data.

Automated sequence annotation is a major goal of post-genomic era with hundreds of genomes in the databases, from both prokaryotes and eukaryotes. While the number of fully sequenced chromosomes from microbial organisms exponentially increased in the last decade above 600, presently we know the whole DNA content of only 25 eukaryotic organisms, including Homo sapiens. However, the process of genome annotation is far from being completed. This is particularly relevant in eukaryotes, whose cells contain several subcellular compartments, or organelles, enclosed by membranes, where different relevant functions are performed. Translocation across the membrane into the organelles is a highly regulated and complex cellular process. Indeed different proteins and/or protein isoforms, originated from genes by alternative splicing, may be conveyed to different cell compartments, depending on their specific role in the cell. During recent years the prediction of subcellular localization (SL) by computational means has been an active research area. Several methods are presently available based on different notions and addressing different aspects of SL. This review provides a short overview of the most well performing methods described in the literature, highlighting their predictive capabilities and different applications. , and works on computational biology and computational biophysics. Pier Luigi Martelli is Researcher at the

Multi-objective genetic local search Self organizing maps Variable Neighborhood Search (VNS) Multi-objective evolutionary algorithm Learning Multi-reservoir operation management a b s t r a c t Genetic Algorithms (GAs) are population based global search methods that can escape from local optima traps and find the global optima regions. However, near the optimum set their intensification process is often inaccurate. This is because the search strategy of GAs is completely probabilistic. With a random search near the optimum sets, there is a small probability to improve current solution. Another drawback of the GAs is genetic drift. The GAs search process is a black box process and no one knows that which region is being searched by the algorithm and it is possible that GAs search only a small region in the feasible space. On the other hand, GAs usually do not use the existing information about the optimality regions in past iterations.

Abstract: The Self-Organizing Map (SOM) represents the result of a vector quantization algorithm that places a number of reference or code-book vectors into a high-dimensional input data space to approximate to its data sets in an ordered fashion. The SOM PAK program package ...

Molecular dynamics simulation methods produce trajectories of atomic positions (and optionally velocities and energies) as a function of time and provide a representation of the sampling of a given molecule's energetically accessible conformational ensemble. As simulations on the 10-100 ns time scale become routine, with sampled configurations stored on the picosecond time scale, such trajectories contain large amounts of data. Data-mining techniques, like clustering, provide one means to group and make sense of the information in the trajectory. In this work, several clustering algorithms were implemented, compared, and utilized to understand MD trajectory data. The development of the algorithms into a freely available C code library, and their application to a simple test example of random (or systematically placed) points in a 2D plane (where the pairwise metric is the distance between points) provide a means to understand the relative performance. Eleven different clustering algorithms were developed, ranging from top-down splitting (hierarchical) and bottom-up aggregating (including single-linkage edge joining, centroid-linkage, average-linkage, complete-linkage, centripetal, and centripetalcomplete) to various refinement (means, Bayesian, and self-organizing maps) and tree (COBWEB) algorithms. Systematic testing in the context of MD simulation of various DNA systems (including DNA single strands and the interaction of a minor groove binding drug DB226 with a DNA hairpin) allows a more direct assessment of the relative merits of the distinct clustering algorithms. Additionally, means to assess the relative performance and differences between the algorithms, to dynamically select the initial cluster count, and to achieve faster data mining by "sieved clustering" were evaluated. Overall, it was found that there is no one perfect "one size fits all" algorithm for clustering MD trajectories and that the results strongly depend on the choice of atoms for the pairwise comparison. Some algorithms tend to produce homogeneously sized clusters, whereas others have a tendency to produce singleton clusters. Issues related to the choice of a pairwise metric, clustering metrics, which atom selection is used for the comparison, and about the relative performance are discussed. Overall, the best performance was observed with the average-linkage, means, and SOM algorithms. If the cluster count is not known in advance, the hierarchical or average-linkage clustering algorithms are recommended. Although these algorithms perform well, it is important to be aware of the limitations or weaknesses of each algorithm, specifically the high sensitivity to outliers with hierarchical, the tendency to generate homogenously sized clusters with means, and the tendency to produce small or singleton clusters with average-linkage.

This paper presents an advanced statistical method for wind power forecasting based on artificial intelligence techniques. The method requires as input past power measurements and meteorological forecasts of wind speed and direction interpolated at the site of the wind farm. A self-organized map is trained to classify the forecasted local wind speed provided by the meteorological services. A unique feature of the method is that following a preliminary wind power prediction, it provides an estimation of the quality of the meteorological forecasts that is subsequently used to improve predictions. The proposed method is suitable for operational planning of power systems with increased wind power penetration, i.e., forecasting horizon of 48 h ahead and for wind farm operators trading in electricity markets. Application of the forecasting method on the power production of an actual wind farm shows the validity of the method.

In this paper, we propose a novel method using wavelets as input to neural network self-organizing maps and support vector machine for classification of magnetic resonance (MR) images of the human brain. The proposed method classifies MR brain images as either normal or abnormal. We have tested the proposed approach using a dataset of 52 MR brain images. Good classification percentage of more than 94% was achieved using the neural network self-organizing maps (SOM) and 98% from support vector machine. We observed that the classification rate is high for a support vector machine classifier compared to self-organizing map-based approach. #

Permission-based security models provide controlled access to various system resources. The expressiveness of the permission set plays an important role in providing the right level of granularity in access control. In this work, we present a methodology for the empirical analysis of permission-based security models which makes novel use of the Self-Organizing Map (SOM) algorithm of . While the proposed methodology may be applicable to a wide range of architectures, we analyze 1,100 Android applications as a case study. Our methodology is of independent interest for visualization of permissionbased systems beyond our present Android-specific empirical analysis. We offer some discussion identifying potential points of improvement for the Android permission model, attempting to increase expressiveness where needed without increasing the total number of permissions or overall complexity.

Self organizing maps (SOMs) are used to locate archetypal points that describe the multi-dimensional distribution function of a gridded sea level pressure data set for the northeast United States. These points -nodes on the SOM -identify the primary features of the synoptic-scale circulation over the region. In effect, the nodes represent a non-linear distribution of overlapping, non-discreet, circulation types. The circulation patterns are readily visualized in a 2-dimensional array (the SOM) that places similar types adjacent to one another and very different types far apart in the SOM space. The SOM is used to describe synoptic circulation changes over time, and to relate the circulation to January station precipitation data (for State College, Pennsylvania) in the center of the domain. The paper focuses on the methodology; however, the analysis suggests that circulation systems that promote precipitation have decreased over the last 40 yr -although January precipitation at State College has actually increased. Further analysis with the SOM indicates that this is due to a change in precipitation characteristics of the synoptic-scale circulation features, rather than to their frequency of occurrence.

In this paper, we propose a novel Intrusion Detection System (IDS) architecture utilizing both anomaly and misuse detection approaches. This hybrid Intrusion Detection System architecture consists of an anomaly detection module, a misuse detection module and a decision support system combining the results of these two detection modules. The proposed anomaly detection module uses a Self-Organizing Map (SOM) structure to model normal behavior. Deviation from the normal behavior is classified as an attack. The proposed misuse detection module uses J.48 decision tree algorithm to classify various types of attacks. The principle interest of this work is to benchmark the performance of the proposed hybrid IDS architecture by using KDD Cup 99 Data Set, the benchmark dataset used by IDS researchers. A rule-based Decision Support System (DSS) is also developed for interpreting the results of both anomaly and misuse detection modules. Simulation results of both anomaly and misuse detection modules based on the KDD 99 Data Set are given. It is observed that the proposed hybrid approach gives better performance over individual approaches.

This paper deals with a new scheme for the prediction of a ball bearing's remaining useful life based on self-organizing map (SOM) and back propagation neural network methods. One of the key components needed for effective bearing life prediction is the set-up of an appropriate degradation indicator from a bearing's incipient defect stage to its final failure. This new method is different from the others that have been used in the past, in that it uses the minimum quantisation error (MQE) indicator derived from SOM, which is trained by six vibration features, including a new designed degradation index for performance degradation assessment. Then, using this indicator, back propagation neural networks focusing on the degradation periods can be trained. Thanks to weight application to failure times (WAFT) technology, a useful life prediction model for ball bearings has been developed successfully. Finally, a set of accelerated bearing run-to-failure experiments is carried out, with the experimental results showing that the new proposed methods are greatly superior to those, based on L10 bearing life prediction, currently being used.

Bearing failure is one of the foremost causes of breakdowns in rotating machinery and such failure can be catastrophic, resulting in costly downtime. One of the key issues in bearing prognostics is to detect the defect at its incipient stage and alert the operator before it develops into a catastrophic failure. Signal de-noising and extraction of the weak signature are crucial to bearing prognostics since the inherent deficiency of the measuring mechanism often introduces a great amount of noise to the signal. In addition, the signature of a defective bearing is spread across a wide frequency band and hence can easily become masked by noise and low frequency effects. As a result, robust methods are needed to provide more evident information for bearing performance assessment and prognostics. This paper introduces enhanced and robust prognostic methods for rolling element bearing including a wavelet filter based method for weak signature enhancement for fault identification and Self Organizing Map (SOM) based method for performance degradation assessment. The experimental results demonstrate that the bearing defects can be detected at an early stage of development when both optimal wavelet filter and SOM method are used.

Abstmct-The self-organizing maps have a bearing on traditional vector quantization. A characteristic that makes them more resemble certain biological brain maps, however, is the spatial order of their responses, which is formed in the learning process. This paper discusses the basic algorithms and two innovations: dynamic weighting of the input signals at each input of each cell, which improves the ordering when very different input signals are used, and definition of neighborhoods in the learning algorithm by the minimal spanning tree, which provides a far better and faster approximation of prominently structured density functions. Finally, it is cautioned that if the maps are used for pattern recognition and decision processes, it is necessary to fine tune the reference vectors so that they directly define the decision borders.

We describe a collection of standardized image processing protocols for electron microscopy singleparticle analysis using the XMIPP software package. These protocols allow performing the entire processing workflow starting from digitized micrographs up to the final refinement and evaluation of 3D models. A particular emphasis has been placed on the treatment of structurally heterogeneous data through maximum-likelihood refinements and self-organizing maps as well as the generation of initial 3D models for such data sets through random conical tilt reconstruction methods. All protocols presented have been implemented as stand-alone, executable python scripts, for which a dedicated graphical user interface has been developed. Thereby, they may provide novice users with a convenient tool to quickly obtain useful results with minimum efforts in learning about the details of this comprehensive package. Examples of applications are presented for a negative stain random conical tilt data set on the hexameric helicase G40P and for a structurally heterogeneous data set on 70S Escherichia coli ribosomes embedded in vitrified ice.

This paper describes a real-valued representation for the negative selection algorithm and its applications to anomaly detection. In many anomaly detection applications, only positive (normal) samples are available for training purpose. However, conventional classification algorithms need samples for all classes (e.g. normal and abnormal) during the training phase. This approach uses only normal samples to generate abnormal samples, which are used as input to a classification algorithm. This hybrid approach is compared against an anomaly detection technique that uses self-organizing maps to cluster the normal data sets (samples). Experiments are performed with different data sets and some results are reported.

The research reported here integrates computational, visual, and cartographic methods to develop a geovisual analytic approach for exploring and understanding spatio-temporal and multivariate patterns. The developed methodology and tools can help analysts investigate complex patterns across multivariate, spatial, and temporal dimensions via clustering, sorting, and visualization. Specifically, the approach involves a self-organizing map, a parallel coordinate plot, several forms of reorderable matrices (including several ordering methods), a geographic small multiple display, and a 2-dimensional cartographic color design method. The coupling among these methods leverages their independent strengths and facilitates a visual exploration of patterns that are difficult to discover otherwise. The visualization system we developed supports overview of complex patterns and, through a variety of interactions, enables users to focus on specific patterns and examine detailed views. We demonstrate the system with an application to the IEEE InfoVis 2005 Contest data set, which contains time-varying, geographically referenced, and multivariate data for technology companies in the US.

Although human epidermal growth factor receptor 2 (HER2) overexpression is implicated in tumor progression for a variety of cancer types, how it dysregulates signaling networks governing cell behavioral functions is poorly understood. To address this problem, we use quantitative mass spectrometry to analyze dynamic effects of HER2 overexpression on phosphotyrosine signaling in human mammary epithelial cells stimulated by epidermal growth factor (EGF) or heregulin (HRG). Data generated from this analysis reveal that EGF stimulation of HER2-overexpressing cells activates multiple signaling pathways to stimulate migration, whereas HRG stimulation of these cells results in amplification of a specific subset of the migration signaling network. Self-organizing map analysis of the phosphoproteomic data set permitted elucidation of network modules differentially regulated in HER2-overexpressing cells in comparison with parental cells for EGF and HRG treatment. Partial least-squares regression analysis of the same data set identified quantitative combinations of signals within the networks that strongly correlate with cell proliferation and migration measured under the same battery of conditions. Combining these modeling approaches enabled association of epidermal growth factor receptor family dimerization to activation of specific phosphorylation sites, which appear to most critically regulate proliferation and/or migration.

Despite its wide applications as a tool for feature extraction, the Self-Organizing Map (SOM) remains a black box to most meteorologists and oceanographers. This paper evaluates the feature extraction performance of the SOM by using artificial data representative of known patterns. The SOM is shown to extract the patterns of a linear progressive sine wave. Sensitivity studies are performed to ascertain the effects of the SOM tunable parameters. By adding random noise to the linear progressive wave data, it is demonstrated that the SOM extracts essential patterns from noisy data. Moreover, the SOM technique successfully chooses among multiple sets of patterns in contrast with an Empirical Orthogonal Function method that fails to do this. A practical way to apply the SOM is proposed and demonstrated using several examples, including long time series of coastal ocean currents from the West Florida Shelf. With improved SOM parameter choices, strong current patterns associated with severe weather forcing are extracted separate from previously identified asymmetric upwelling/downwelling and transitional patterns associated with more typical weather forcing.

Salt stress is one of the most important factors limiting plant cultivation. Many investigations of plant response to high salinity have been performed using conventional transcriptomics and/or proteomics approaches. However, transcriptomics and proteomics techniques are not all-encompassing methods that can achieve exclusive insights into the metabolite networks contributing to biochemical reactions. Hence, the functions of the complex stress response pathways are yet to be determined, especially at the metabolic level. A timecourse metabolic profiling with Arabidopsis thaliana cell cultures after the imposition of salt stress is reported in this study. Analyses of primary metabolites, especially small polar metabolites such as amino acids, sugars, sugar alcohols, organic acids, and amines, was performed by GC/MS and LC/MS at 0.5, 1, 2, 4, 12, 24, 48, and 72 h after a salt-stress treatment with 100 mM NaCl being the final concentration. The mass chromatographic data were converted into matrix data sets, which were subjected to data mining processes, including principal component analysis (PCA) and batchlearning self-organizing mapping analysis (BL-SOM). The mining results suggest that the methylation cycle for the supply of methyl groups, the phenylpropanoid pathway for lignin production, and glycinebetaine biosynthesis are synergetically induced as a short-term response against salt-stress treatment. The results also suggest the the co-induction of glycolysis and sucrose metabolism as well as co-reduction of the methylation cycle as long-term responses to salt stress.

An image segmentation system is proposed for the segmentation of color image based on neural networks. In order to measure the color difference properly, image colors are represented in a modified color space. The segmentation system comprises unsupervised segmentation and supervised segmentation. The unsupervised segmentation is achieved by a two-level approach, i.e., color reduction and color clustering. In color reduction, image colors are projected into a small set of prototypes using self-organizing map (SOM) learning. In color clustering, simulated annealing (SA) seeks the optimal clusters from SOM prototypes. This two-level approach takes the advantages of SOM and SA, which can achieve the near-optimal segmentation with a low computational cost. The supervised segmentation involves color learning and pixel classification. In color learning, color prototype is defined to represent a spherical region in color space. A procedure of hierarchical prototype learning (HPL) is used to generate the different sizes of color prototypes from the sample of object colors. These color prototypes provide a good estimate for object colors. The image pixels are classified by the matching of color prototypes. The experimental results show that the system has the desired ability for the segmentation of color image in a variety of vision tasks.

Development of content-based image retrieval (CBIR) techniques has suffered from the lack of standardized ways for describing visual image content. Luckily, the MPEG-7, or formally "Moving Pictures Expert Group Multimedia Content Description Interface" international standard is now emerging as both a general framework for content description and a collection of specific agreed-upon content descriptors. We have developed a neural, self-organizing technique for CBIR. Our system is named PicSOM and it is based on pictorial examples and relevance feedback (RF). The name stems from "picture" and the self-organizing map (SOM). The PicSOM system is implemented by using tree structured SOMs. In this paper, we apply the visual content descriptors provided by MPEG-7 in the PicSOM system and compare our own image indexing technique with a reference system based on vector quantization (VQ). The results of our experiments show that the MPEG-7-defined content descriptors can be used as such in the PicSOM system even though Euclidean distance calculation, inherently used in the PicSOM system, is not optimal for all of them. Also, the results indicate that the PicSOM technique is a bit slower than the reference system in starting to find relevant images. However, when the strong RF mechanism of PicSOM begins to function, its retrieval precision exceeds that of the reference system. Index Terms-Content-based image retrieval (CBIR), MPEG-7, query by pictorial example (QBPE), relevance feedback (RF), selforganizing map (SOM), visual content description.

In this paper we present a comparison among some nonhierarchical and hierarchical clustering algorithms including SOM (Self-Organization Map) neural network and Fuzzy c-means methods. Data were simulated considering correlated and uncorrelated variables, nonoverlapping and overlapping clusters with and without outliers. A total of 2530 data sets were simulated. The results showed that Fuzzy c-means had a very good performance in all cases being very stable even in the presence of outliers and overlapping. All other clustering algorithms were very affected by the amount of overlapping and outliers. SOM neural network did not perform well in almost all cases being very affected by the number of variables and clusters. The traditional hierarchical clustering and K-means methods presented similar performance.

By slightly changing the de nition of the winning unit, Kohonen's original learning rule can be viewed as performing stochastic gradient descent on an energy function. We show this in two ways: by explicitely computing derivatives and as a limiting case of a \soft" version of self-organizing maps with probabilistic winner assignments. Kinks in a one-dimensional map and twists in a two-dimensional map correspond to local minima in the energy landscape of the network weights.

Digital image libraries are becoming more common and widely used as more visual information is produced at a rapidly growing rate. Content-based image retrieval is an important approach to the problem of processing this increasing amount of data. It is based on automatically extracted features from the content of the images, such as color, texture, shape, and structure. We have started a project to study methods for content-based image retrieval using the Self-Organizing Map (SOM) as the image similarity scoring method. Our image retrieval system, named Pic-SOM, can be seen as a SOM-based approach to relevance feedback which is a form of supervised learning to adjust the subsequent queries based on the user's responses during the information retrieval session. In PicSOM, a separate Tree Structured SOM (TS-SOM) is trained for each feature vector type in use. The system then adapts to the user's preferences by returning her more images from those SOMs where her responses have been most densely mapped.

In the paper very efficient, linear in number of arcs, algorithms for determining Hummon and Doreian's arc weights SPLC and SPNP in citation network are proposed, and some theoretical properties of these weights are presented. The nonacyclicity problem in citation networks is discussed. An approach to identify on the basis of arc weights an important small subnetwork is proposed and illustrated on the citation networks of SOM (self organizing maps) literature and US patents.

The paper presents a new framework for the classification of polarimetric SAR data. The underlying model introduces cyclic conditional dependencies among the class labels assigned to neighboring observations as a mechanism to regulate the spatial homogeneity of classification results. Classification is posed as an inference problem, and is solved by coherently integrating expectation maximization and graph cut optimization. Results based on real SAR data are presented.

Information visualization is an interdisciplinary research area in which cartographic efforts have mostly addressed the handling of geographic information. Some cartographers have recently become involved in attempts to extend geographic principles and cartographic techniques to the visualization of non-geographic information. This paper reports on current progress and future opportunities in this emerging research field commonly known as spatialization. The discussion is mainly devoted to the computational techniques that turn high-dimensional data into visualizations via processes of projection and transformation. It is argued that cartographically informed engagement of computationally intensive techniques can help to provide richer and less opaque information visualizations. The discussion of spatialization methods is linked to another priority area of cartographic involvement, the development of theory and principles for cognitively plausible spatialization. The paper distinguishes two equally important sets of challenges for cartographic success in spatialization research. One is the recognition that there are distinct advantages to applying a cartographic perspective in information visualization. This requires our community to more thoroughly understand the essence of cartographic activity and to explore the implications of its metaphoric transfer to non-geographic domains. Another challenge lies in cartographers becoming a more integral part of the information visualization community and actively engaging its constituent research fields.

The largest damming project to date, the Three Gorges Dam has been built along the Yangtze River (China), the most species-rich river in the Palearctic region. Among 162 species of fish inhabiting the main channel of the upper Yangtze, 44 are endemic and are therefore under serious threat of global extinction from the dam. Accordingly, it is urgently necessary to develop strategies to minimize the impacts of the drastic environmental changes associated with the dam. We sought to identify potential reserves for the endemic species among the 17 tributaries in the upper Yangtze, based on presence/absence data for the 44 endemic species. Potential reserves for the endemic species were identified by characterizing the distribution patterns of endemic species with an adaptive learning algorithm called a "self-organizing map" (SOM). Using this method, we also predicted occurrence probabilities of species in potential reserves based on the distribution patterns of communities. Considering both SOM model results and actual knowledge of the biology of the considered species, our results suggested that 24 species may survive in the tributaries, 14 have an uncertain future, and 6 have a high probability of becoming extinct after dam filling.

The soil sorption partition coefficient (log K oc ) of a heterogeneous set of 643 organic non-ionic compounds, with a range of more than 6 log units, is predicted by a statistically validated QSAR modeling approach. The applied multiple linear regression (ordinary least squares, OLS) is based on a variety of theoretical molecular descriptors selected by the genetic algorithms-variable subset selection (GA-VSS) procedure. The models were validated for predictivity by different internal and external validation approaches. For external validation we applied self organizing maps (SOM) to split the original data set: the best four-dimensional model, developed on a reduced training set of 93 chemicals, has a predictivity of 78% when applied on 550 validation chemicals (prediction set). The selected molecular descriptors, which could be interpreted through their mechanistic meaning, were compared with the more common physico-chemical descriptors log K ow and log S w . The chemical applicability domain of each model was verified by the leverage approach in order to propose only reliable data. The best predicted data were obtained by consensus modeling from 10 different models in the genetic algorithm model population. #

The Self-Organizing Map (SOM) is a powerful neural network method for analysis and visualization of high-dimensional data. It maps nonlinear statistical dependencies between high-dimensional measurement data into simple geometric relationships on a usually twodimensional grid. The mapping roughly preserves the most important topological and metric relationships of the original data elements and, thus, inherently clusters the data. The need for visualization and clustering occur, for instance, in the analysis of various engineering problems. In this paper, the SOM has been applied in monitoring and modeling of complex industrial processes. Case studies, including pulp process, steel production, and paper industry are described.

Motivation: Recent advancements in microarray technology allows simultaneous monitoring of the expression levels of a large number of genes over different time points. Clustering is an important tool for analyzing such microarray data, typical properties of which are its inherent uncertainty, noise and imprecision. In this article, a two-stage clustering algorithm, which employs a recently proposed variable string length genetic scheme and a multiobjective genetic clustering algorithm, is proposed. It is based on the novel concept of points having significant membership to multiple classes. An iterated version of the well-known Fuzzy C-Means is also utilized for clustering. Results: The significant superiority of the proposed two-stage clustering algorithm as compared to the average linkage method, Self Organizing Map (SOM) and a recently developed weighted Chinese restaurant-based clustering method (CRC), widely used methods for clustering gene expression data, is established on a variety of artificial and publicly available real life data sets. The biological relevance of the clustering solutions are also analyzed.

Metastasis is a complicated multistep process that involves interactions between cancer cells and their surrounding microenvironments. Previously, we have established a series of lung adenocarcinoma cell lines with varying degrees of invasiveness. Tracheal graft assay confirmed that cell lines with higher in vitro invasiveness had greater in vivo invasive potential. In this study, we used these model cell lines to identify invasion-associated genes using cDNA microarray with colorimetric detection. A more invasive subline, CL 1-5-F 4, derived from metastatic lung tumor of severe combined immunodeficient mice inoculated with CL 1-5 cells, was combined with CL 1-0, CL 1-1, and CL 1-5 in cDNA microarray screening. cDNA microarray membranes, each containing 9600 nonredundant expressed sequence tag clones, were used to identify differentially expressed genes in these cell lines. For statistical analysis, self-organizing map algorithm was performed to identify the expression patterns. Pos...