|
University of Jyväskylä
Department of Mathematical Information Technology
NDA - Neural Data Analysis environment
|
Neural Data Analysis (NDA)
A generally applicable software kernel for data analysis and data mining tasks.
- The NDA provides a large set of methods for data manipulation, data reduction, data visualization and decision making.
- The most important algorithms in the software are the Tree Structured Self-Organizing Map, the Multilayer Perceptron network with the mostly used variants of the backpropagation algorithm, multivariate statistics, and a lot of data processing methods.
- Please navigate the software through the figure:
- The NDA supports best the interactive SOM analysis. See an example.
The Software Architecture
The NDA is a portable software kernel.
- It is independent of used operating systems and graphical environments:
It is easy to be ported for different operating systems (Windows, Unix).
It can be used with different development environments (Java, Motif, Excel, Visual Basic,...).
- The kernel can be easily tailored for different applications.
See more about the architecture of the NDA.
Applications and examples
NDA is used either as a data analysis tool or as a software kernel in applications. Our collaborators have built their own software products based on the NDA kernel. Some examples are
- Process monitoring and control, analysis of process data (Visipoint Ltd, Pechiney ElectroMetallurgy)
- Quality control of the print paper (Hormell Ltd)
- Detection of similar text documents (Sonera Ltd)
- Analysis of qualitative data (CATO project)
- Decision support in logistics planning (UPM-Kymmene Logistics Ltd)
- Air quality control and prediction (Laboratory of Environmental Chemistry)
- Analysis and visualization of gene expressions (Visipoint Ltd)
Documents
User's command reference: User's Guide.
(Latest old version of User's Guide.)
(Old version of User's Guide.)
(Even older version of User's Guide.)
(Ancient version of User's Guide.)
Distribution
Index for distributed files (password needed)
Back to SECI homepage
This page was last modified on May 31st, 2000 by Erkki Häkkinen