Publisher's Description
From Plum Amazing:
KnowledgeMiner is a data mining tool that enables anyone to use its unique form of modeling to quickly visualize new possibilities. It is an artificial intelligence tool designed to easily extract hidden knowledge from data. It was built on the cybernetic principles of self-organization: Learning a completely unknown relationship between output and input of any given system in an evolutionary way from a very simple organization to an optimally complex one.
One of KnowledgeMiner's major strengths is that it extensively supports prediction of dynamic systems. You can use GMDH, Fuzzy Rule Induction, and Analog Complexing to predict the behavior of any time-based process. Data mining is widely considered to deal only with static systems. Therefore, the majority of data mining tools only support or focus on static systems, such as classification, segmentation and clustering.
Another advantage of KnowlegeMiner is its knowledge extraction capabilities. It self-selects some relevant inputs, generates an optimal composition (model), and provides an explanation component (equation, rule, or patterns) on the fly.
Please see the website at knowledgeminer.com for a more information and examples.
What's new in this version:
- the package that allows knowledgeminer to run under ms windows is now free.
- this means everything is now inlcuded to run knowledgeminer under windows.
- reports on the fly a model's inputs' contribution to increasing model accuracy;
- optionally, for prediction, checks and limits an input variable's value if it is within the range used for building the model (range check);
- on the fly calculation of the coefficient of determination (R-squared) and the Me... See all new features
More Popular Math Software downloads
- IBM SPSS Statistics
1,930 downloads
- Matlab
1,395 downloads
- MathSoft Mathcad PLUS 6
291 downloads
- Stata
266 downloads
- MathType
210 downloads
Add Your Review
Submit your reply
E-mail this review
Report offensive content
See more CNET content tagged:
Previous Versions: