KDnuggets : News : 2008 : n10 : item30 < PREVIOUS | NEXT >


Health Discovery Corporation Announces the Issuance of Two New U.S. Patents Covering Applications of Support Vector Machines (SVM's)

News source: Business Wire

Health Discovery Corporation (OTCBB: HDVY) today announced that the U.S. Patent and Trademark Office has issued two new patents to further expand HDC's exclusive intellectual property holdings covering a variety of applications of SVMs to a wide range of technologies. The first patent, Patent No. 7,353,215, claims a method for analysis of any type of data that has a structure. One of the most familiar types of structured data is documents, making the patent claims highly relevant to commercial data mining applications, including Internet browsers and search engines.

Other commercial applications of this newly issued patent include document management and sorting, text comparison, plagiarism and forgery detection, handwriting recognition and signature verification, biometrics, and any number of other applications in the high technology arena. Other types of structured data covered by this patent include images and electromagnetic signals, such as wireless transmissions and spectra. Information extraction and pattern recognition capabilities within such structured data are widely being pursued by industries including communications, manufacturing, national defense, and oil exploration.


The second issued patent, Patent No. 7,318,051, covers additional feature selection techniques that can be used to successfully identify the most important pieces of information needed to solve complex pattern-recognition problems. HDC's previously patented recursive feature elimination method (RFE-SVM) is a type of feature selection technique that has proven to be successful in a wide variety of applications. These feature selection methods are critical components often used as an important "pre-processing" step for the successful implementation of support vector machine (SVM) technology.

Read more.

Bookmark using any bookmark manager!

KDnuggets : News : 2008 : n10 : item30 < PREVIOUS | NEXT >

Copyright © 2008 KDnuggets.   Subscribe to KDnuggets News!