Advances in Self-Organizing Maps and Learning Vector by Thomas Villmann, Frank-Michael Schleif, Marika Kaden, Mandy

By Thomas Villmann, Frank-Michael Schleif, Marika Kaden, Mandy Lange (eds.)

The booklet collects the medical contributions offered on the tenth Workshop on Self-Organizing Maps (WSOM 2014) held on the collage of technologies Mittweida, Mittweida (Germany, Saxony), on July 2–4, 2014. beginning with the 1st WSOM-workshop 1997 in Helsinki this workshop makes a speciality of most up-to-date ends up in the sphere of supervised and unsupervised vector quantization like self-organizing maps for facts mining and information classification.

This tenth WSOM introduced jointly greater than 50 researchers, specialists and practitioners within the appealing small city Mittweida in Saxony (Germany) within sight the mountains Erzgebirge to debate new advancements within the box of unsupervised self-organizing vector quantization platforms and studying vector quantization methods for category. The ebook includes the approved papers of the workshop after a cautious evaluate strategy in addition to summaries of the invited talks. between those e-book chapters there are very good examples of using self-organizing maps in agriculture, desktop technology, info visualization, well-being structures, economics, engineering, social sciences, textual content and snapshot research and time sequence research. different chapters current the newest theoretical paintings on self-organizing maps in addition to studying vector quantization tools, reminiscent of referring to these easy methods to classical statistical selection methods.

All the contribution exhibit that vector quantization tools hide a wide range of software parts together with facts visualization of high-dimensional complicated facts, complex selection making and type or info clustering and knowledge compression.

Show description

By Thomas Villmann, Frank-Michael Schleif, Marika Kaden, Mandy Lange (eds.)

The booklet collects the medical contributions offered on the tenth Workshop on Self-Organizing Maps (WSOM 2014) held on the collage of technologies Mittweida, Mittweida (Germany, Saxony), on July 2–4, 2014. beginning with the 1st WSOM-workshop 1997 in Helsinki this workshop makes a speciality of most up-to-date ends up in the sphere of supervised and unsupervised vector quantization like self-organizing maps for facts mining and information classification.

This tenth WSOM introduced jointly greater than 50 researchers, specialists and practitioners within the appealing small city Mittweida in Saxony (Germany) within sight the mountains Erzgebirge to debate new advancements within the box of unsupervised self-organizing vector quantization platforms and studying vector quantization methods for category. The ebook includes the approved papers of the workshop after a cautious evaluate strategy in addition to summaries of the invited talks. between those e-book chapters there are very good examples of using self-organizing maps in agriculture, desktop technology, info visualization, well-being structures, economics, engineering, social sciences, textual content and snapshot research and time sequence research. different chapters current the newest theoretical paintings on self-organizing maps in addition to studying vector quantization tools, reminiscent of referring to these easy methods to classical statistical selection methods.

All the contribution exhibit that vector quantization tools hide a wide range of software parts together with facts visualization of high-dimensional complicated facts, complex selection making and type or info clustering and knowledge compression.

Show description

Read Online or Download Advances in Self-Organizing Maps and Learning Vector Quantization: Proceedings of the 10th International Workshop, WSOM 2014, Mittweida, Germany, July, 2-4, 2014 PDF

Best international_1 books

CONCUR 2001 — Concurrency Theory: 12th International Conference Aalborg, Denmark, August 20–25, 2001 Proceedings

This booklet constitutes the refereed complaints of the twelfth overseas convention on Concurrency conception, CONCUR 2001, held in Aalborg, Denmark in August 2001. The 32 revised complete papers offered including six invited contributions have been conscientiously reviewed and chosen from seventy eight submissions. The papers are prepared in topical sections on mobility, probabilistic platforms, version checking, procedure algebra, unfoldings and prefixes, good judgment and compositionality, and video games.

E-Business and Telecommunications: International Joint Conference, ICETE 2013, Reykjavik, Iceland, July 29-31, 2013, Revised Selected Papers

This booklet constitutes the refereed complaints of the tenth foreign Joint convention on E-Business and Telecommunications, ICETE 2013, held in Reykjavik, Iceland, in July 2013. ICETE is a joint overseas convention integrating 4 significant components of data which are divided into six corresponding meetings: overseas convention on facts communique Networking, DCNET; foreign convention on E-Business, ICE-B; overseas convention on Optical communique platforms, OPTICS; overseas convention on protection and Cryptography, SECRYPT; overseas convention on instant details platforms, WINSYS; and overseas convention on sign Processing and Multimedia, SIGMAP.

Quantum Interaction: 10th International Conference, QI 2016, San Francisco, CA, USA, July 20-22, 2016, Revised Selected Papers

This publication constitutes the completely refereed post-conference complaints of the tenth foreign convention on Quantum interplay, QI 2016, held in San Francisco, CA, united states, in July 2016. The 21 papers offered during this e-book have been conscientiously reviewed and chosen from 39 submissions. The papers deal with subject matters similar to: basics; Quantum Cognition; Language and functions; Contextuality and Foundations of chance; and Quantum-Like Measurements.

Pattern Recognition Applications and Methods: 5th International Conference, ICPRAM 2016, Rome, Italy, February 24-26, 2016, Revised Selected Papers

This ebook comprises revised and prolonged types of chosen papers from the fifth foreign convention on development popularity, ICPRAM 2016, held in Rome, Italy, in February 2016. The thirteen complete papers have been conscientiously reviewed and chosen from a hundred twenty five preliminary submissions and describe up to date purposes of development popularity options to real-world difficulties, interdisciplinary examine, experimental and/or theoretical reports yielding new insights that enhance development acceptance tools.

Extra info for Advances in Self-Organizing Maps and Learning Vector Quantization: Proceedings of the 10th International Workshop, WSOM 2014, Mittweida, Germany, July, 2-4, 2014

Sample text

2 A Self-organizing Dynamic Neural Field We now embed the neural field equation within a learning architecture similar to Kohonen Self-Organizing Maps. Indeed, we consider the dynamical system given by Eq. (4). N ] in the neural field. At regular time intervals, a sample ζ, drawn from an unknown distribution D ⊆ R2 , is presented to the field. The input I(x, t) at each position is computed as a Gaussian, with standard deviation σ, of ζ centered on p(x, t). Importantly, the prototypes are updated at every iteration and not just when the field has converged.

Distance to ordinate axis : M F2 (i, X ) = abs(wi2 ). Distance to point (0,0): M F3 (i, X ) = wi . Distance to the mean of dataset: M F4 (i, X ) = wi − x mean . Figure 1 shows some results for SOM and MS-SOM (with 80 units) of the grid representation over the data space. Figure 1(a) shows the corresponding D-matrix (mean distances from weights of neighbour units), and the magnitude map for one MS-SOM. Figure 1(b) shows the typical result of a trained SOM where units tend to allocate their centroids in areas with higher data density.

Karlsruhe (1994) 9. fr Abstract. In a number of real-life applications, the user is interested in analyzing non vectorial data, for which kernels are useful tools that embed data into an (implicit) Euclidean space. However, when using such approaches with prototype-based methods, the computational time is related to the number of observations (because the prototypes are expressed as convex combinations of the original data). Also, a side effect of the method is that the interpretability of the prototypes is lost.

Download PDF sample

Rated 4.15 of 5 – based on 50 votes