Self-Organizing Maps

Teuvo Kohonen (2001) Springer Verlag, Third Edition

This is the basic monography and standard lecture book by Teuvo Kohonen, the originator of the Self-Organizing Maps, about that subject.

SOM, with its variants, is the most popular artificial neural network algorithm in the unsupervised learning category. Many fields of science, including statistics, signal processing, control theory, financial analyses, experimental physics, chemistry and medicine, have adopted the SOM as a standard analytical tool.

Viscovery SOMine is reviewed as a commercial standard tool for creating SOM applications.

Kohonen Maps

Erkki Oja, Samuel Kaski; Contributor Erkki Oja, Samuel Kaski (1999), Elsevier

Top experts in the SOM method cover the state of the art and the future of theory. The 30 chapters of this book cover the current status of SOM theory, such as connections of SOM to clustering, classification, probabilistic models, and energy functions. Many applications of the SOM are given, with data mining and exploratory data analysis, the central topic, applied to large databases of financial data, medical data, free-form text documents, digital images, speech, and process measurements. Biological models related to the SOM are discussed as well.

Evolutionary multiobjective optimization and visualization

Obayashi (2005) In: K Kuwahara, K Fujii, MPK Fujii, eds. New Developments in Computational Fluid Dynamics: Proceedings of the Sixth International Nobeyama Workshop on the New Century of Computational Fluid Dynamics, Nobeyama, Japan, April 21 to 24, 2003 Springer Verlag

This volume contains 20 papers presented at the Sixth International Nobeyama Workshop on the New Century of Computational Fluid Dynamics, Nobeyama, Japan, April 21-24, 2003.

Viscovery SOMine was used to cluster the parameter space of supersonic wing design.

Unlocking successful new rural industries: Is supply chain management the key?

Collins and Lim-Camacho (2005) Rural Industries Research and Development Corporation (Publication No 05/139)

This final report to the Rural Industries Research and Development Corporation addresses the role of supply chain management as a business strategy in the development of five new rural industries in Australia. It demonstrates that in new rural industries, firms with medium to high levels of supply chain management thinking, planning, and implementation are more successful in terms of creating value and achieving competitive advantage than those with lower levels of supply chain management.

Viscovery SOMine, SPSS, and qualitative analysis were used to elucidate patterns across different supply chain case studies. Clustering revealed how unique characteristics of individual cases combine to exhibit distinct patterns of similarities and differences.

Tradeoff analysis of aerodynamic wing design for RLV

Chiba and Obayashi (2004) In: Parallel Computational Fluid Dynamics 2004: Multidisciplinary Applications G Winter, J Periaux, A Ecer, N Satofuka, P Fox eds. Elsevier

This book reports proceedings from Parallel CFD 2004, the sixteenth international conference on Parallel Computational Fluid Dynamics and other modern scientific domains (May 2004). The sessions involved contributed papers on many diverse subjects including turbulence, complex flows, unstructured and adaptive grids, industrial applications, developments in software tools and environments as parallel optimization tools.

Viscovery SOMine was used as a data mining technique to acquire the knowledge about design space. The obtained results were considered useful for solving optimization problems.

Applications of Multi-Objective Evolutionary Algorithms

Coello and Lamont (2004) World Scientific

Detailed MOEA applications are discussed by international experts, covering state-of-the-art practical insights in tackling statistical optimization with MOEAs.

Viscovery SOMine was used for cluster analysis of design tradeoffs and design variables. The analysis uncovered important relationships and demonstrated the use of the software for engineering design.

Soft systems analysis of ecosystems

Shanmuganathan (2004) PhD thesis, University of Technology, Auckland

This thesis describes case study evaluation of the use of SOM techniques for ecosystem modeling to overcome the perceived inadequacies with conventional ecological data analysis methods.

Viscovery SOMine was used for SOM-based clustering for financial data analysis. The approach proved to be an excellent tool for analyzing the changes within physical system variables and their effects on the biological systems analyzed.

Neural Networks in Business: Techniques and Applications

KA Smith and JND Gupta, eds. (2003) PA: Ideas Group Inc.

This book briefly reviews the two main types of neural networks that find application in business: multilayered feedforward neural networks (MFNN) and self-organizing maps (SOM). Several use cases for these methods are discussed and guidelines are provided for facilitating the successful application of these neural network models to business problems.

Viscovery SOMine software was used for SOM training of countries based on their economic performance. Critical guidelines for facilitating the successful application of neural network methods to business problems are provided.

Features, objects and other things: ontological distinctions in the geographical domain

Mark, Skupin and Smith (2001) In: DR Montello, ed. Spatial Information Theory: Foundations of Geographic Information Science : International Conference, COSIT 2001, Morro Bay, CA, USA, September 19-23, 2001 : Proceedings, Springer Verlag

This book constitutes the refereed proceedings of the 5th International Conference on Spatial Information Theory, COSIT 3001, held in Morro Bay, CA, USA in September 2001.

Viscovery SOMine was used for SOM training and initial labeling of neurons to uncover relationships in subject responses to elicitation questions designed to further a geospatial ontology research project.

Visual Explorations in Finance: With Self-Organizing Maps

Deboeck and Kohonen (1998) Springer Verlag

This book looks at near state-of-the-art SOM applications in finance, economic, and marketing applications. The book contains chapters on applications of unsupervised neural networks using Kohonen's self-organizing map approach to consumer attitudes in Beijing and Shanghai.

The application of SOMs to the modeling and analyses of financial data is demonstrated in detail using Viscovery SOMine. Sample data sets on mutual funds, emerging markets, and responses to a CEIBS questionnaire are used to analyze investment opportunities and credit risk and to uncover consumer attitudes in Beijing and Shanghai.

Neuronale Netze im Marketing-Management: Praxisorientierte Einführung in modernes Data-Mining (German language)

Klaus-Peter Wiedmann, Frank Buckler (2003) Gabler Verlag

This book presents software solutions for the application of neural networks to marketing management. A comprehensive overview of all relevant methods, software, and software vendors is provided.

Viscovery SOMine was evaluated and found to be an excellent tool for analyzing CBM data, including relationships in large data sets, representing prediction models, and creating customer profiles.

Prognose der Kreditausfallswahrscheinlichkeit. Mit künstlichen neuronalen Netzen. (German language)

Andreas T. Moser (2010) VDM Verlag

This book focuses on the development of models that are able to forecast insolvency. A method is developed that finds a quantitative indicator by analyzing and testing huge sets of nonlinear models and filtering the most effective models.

Viscovery Profiler was used to visualize key figures and the scores on them and to derive profiles of the effects of these key figures.