发明名称 |
Methods and systems for feature extraction |
摘要 |
A method and system for extracting feature utilizing an AHaH module (Anti-Hebbian and Hebbian). A sparse input data stream can be presented to a synaptic matrix of a collection of AHaH nodes associated with the AHaH module. The AHaH module operates an AHaH plasticity rule via an evaluate phase and a feedback phase cycle. A bias input line can be modulated such that a bias weight do not receive a Hebbian portion of the weight update during the feedback phase in order to prevent occupation of a null state. The input space can be bifurcated when the AHaH nodes fall randomly into an attractor state. The output of the AHaH module that forms a stable bit pattern can then be provided as an input to a content-addressable memory for generating a maximally efficient binary label. |
申请公布号 |
US8918353(B2) |
申请公布日期 |
2014.12.23 |
申请号 |
US201213595158 |
申请日期 |
2012.08.27 |
申请人 |
KnowmTech, LLC |
发明人 |
Nugent Alex |
分类号 |
G06F15/18;G06N3/08;G06N99/00;G06N5/02 |
主分类号 |
G06F15/18 |
代理机构 |
Ortiz & Lopez, PLLC |
代理人 |
Lopez Kermit D.;Ortiz Luis M.;Ortiz & Lopez, PLLC |
主权项 |
1. A feature extraction method, said method comprising:
presenting an input data stream to a synaptic matrix of a plurality of Anti-Hebbian and Hebbian nodes associated with an Anti-Hebbian and Hebbian module that operates a plasticity rule via an evaluate phase cycle and a feedback phase cycle; modulating a bias input line such that a bias weight does not receive a Hebbian portion of a weight update during said feedback phase in order to prevent occupation of a null state, which does not partition Anti-Hebbian and Hebbian inputs; partitioning an input space with respect to said input data stream when said plurality of Anti-Hebbian and Hebbian nodes falls randomly into an attractor state; and providing an output of said Anti-Hebbian and Hebbian module that forms a stable bit pattern as an input to a content-addressable memory for generating a maximally efficient binary label. |
地址 |
Albuquerque NM US |