Pattern discovery is a new methodology in the armamentarium
of empirical modeling. It describes a family of methods
within the category of knowledge discovery and data mining
that finds relationships among complex, high-dimensionality
data as it relates to underlying information content and
semantic meaning, without the need for an a Cira has developed novel approaches to pattern discovery that can be applied to cytomic, genomic, and proteomic data. Importantly, these methods can be used to integrate different data types such as those found in chemistry and biology. Patent applications for these pattern discovery methods are in preparation.
In a simple data set, like the one derived from For very high dimensionality data this poses a severe challenge for any analytic method. As the dimensionality grows, the number of possible interactions among variables increases exponentially. Cira's pattern discovery methods offer a powerful approach that is capable of discovering and ranking all possible interactions exhaustively. Experience shows that the method is polynomial in time and space, even though the problem is exponential in principle. Cira's methodology is therefore capable of discovering solutions that, by their nature, are hidden from other algorithmic approaches. In addition, our technology is inherently capable of bridging multiple data domains, ensuring its ability to access mutual information available by considering all of the data in a holistic framework. The result is that our technologies offer the opportunity to gain unique insights into your data and to derive unique competitive advantage from it. |