Back to results
US 2014/0006323 A1Published application
Extensible adaptive classification framework
AssigneeKnowmTech, LLC
InventorAlex Nugent
Life of this patent
Estimating the term…
Abstract
An extensible adaptive classification framework and method can include multiple feature detection modules, and a platform for integrating the multiple feature detection modules utilizing a plurality of AHaH nodes as adaptive classifiers over a feature space of multiple and extensible feature factory modules, thereby configuring the platform as an extensible and continuously adaptive pattern recognition platform.
Claims20
An extensible adaptive classification framework, comprising:
- multiple and extensible feature factory modules;
- a plurality of AHaH nodes operating as adaptive classifiers over a space of said multiple an extensible feature factory modules, wherein said multiple and extensible feature factor modules and said plurality of AHaH nodes together provide an extensible and continuously adaptive pattern recognition platform.
An extensible adaptive classification framework, comprising:
- multiple and extensible feature factory modules;
- a plurality of AHaH nodes operating as adaptive classifiers over a space of said multiple an extensible feature factory modules, wherein said multiple and extensible feature factor modules and said plurality of AHaH nodes together provide an extensible and continuously adaptive pattern recognition platform, wherein each feature factory module among said multiple and extensible factor modules employs at least one weight therebetween and at least one AHaH node among said plurality of AHaH nodes as a signal for self-optimization.
An extensible adaptive classification method, comprising:
- configuring a multiple and extensible feature factory modules;
- and
- providing a plurality of AHaH nodes that operate as adaptive classifiers over a space of said multiple an extensible feature factory modules, wherein said multiple and extensible feature factor modules and said plurality of AHaH nodes together provide an extensible and continuously adaptive pattern recognition platform.
Dependencies are inferred from the claim text.