US 11,875,784 B2
Methods and systems for optimized selection of data features for a neuro-linguistic cognitive artificial intelligence system
Gang Xu, Katy, TX (US); Tao Yang, Katy, TX (US); and Ming-Jung Seow, The Woodlands, TX (US)
Assigned to Intellective Ai, Inc., Dallas, TX (US)
Filed by Intellective Ai, Inc., Addison, TX (US)
Filed on Nov. 30, 2020, as Appl. No. 17/107,383.
Application 17/107,383 is a continuation of application No. 16/120,943, filed on Sep. 4, 2018, granted, now 10,853,661.
Application 16/120,943 is a continuation in part of application No. 15/481,298, filed on Apr. 6, 2017, abandoned.
Application 15/481,298 is a continuation in part of application No. PCT/US2017/026419, filed on Apr. 6, 2017.
Claims priority of provisional application 62/318,999, filed on Apr. 6, 2016.
Claims priority of provisional application 62/319,170, filed on Apr. 6, 2016.
Prior Publication US 2021/0165958 A1, Jun. 3, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 15/16 (2006.01); G10L 15/197 (2013.01); G06N 20/00 (2019.01); G06F 40/237 (2020.01); G06V 20/52 (2022.01); G06V 40/20 (2022.01)
CPC G10L 15/16 (2013.01) [G06F 40/237 (2020.01); G06N 20/00 (2019.01); G06V 20/52 (2022.01); G06V 40/20 (2022.01); G10L 15/197 (2013.01)] 23 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, via a processor and from a sensor, a first vector of input data, the first vector of input data indicating a type of the sensor;
identifying, via the processor, the sensor based on the first vector of input data;
generating, via the processor and based on a feature-combination rule set, a plurality of feature symbols by organizing the first vector of input data into probabilistic clusters;
generating, via the processor, an adaptive linguistic model based at least in part on a statistical distribution of the plurality of feature symbols and the type of the sensor;
repeatedly updating the adaptive linguistic model based on subsequent received vectors of input data; and
determining, based on the type of the sensor, a plurality of optimization parameters to optimize feature selection.