US 11,657,270 B2
Self-assessing deep representational units
Savitha Ramasamy, Singapore (SG); and Rajaraman Kanagasabai, Singapore (SG)
Assigned to Agency for Science, Technology and Research, Singapore (SG)
Appl. No. 16/651,637
Filed by Agency for Science, Technology and Research, Singapore (SG)
PCT Filed Sep. 28, 2017, PCT No. PCT/SG2017/050486
§ 371(c)(1), (2) Date Mar. 27, 2020,
PCT Pub. No. WO2019/066718, PCT Pub. Date Apr. 4, 2019.
Prior Publication US 2020/0311544 A1, Oct. 1, 2020
Int. Cl. G06N 3/08 (2023.01); G06N 3/04 (2023.01)
CPC G06N 3/08 (2013.01) [G06N 3/04 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A method of feature learning carried out by a computing device, the method comprising:
receiving a data sample as an input to a feature learning model;
calculating a reconstruction error based on the data sample and a plurality of features of the feature learning model;
determining whether the reconstruction error satisfies a first threshold;
adding a feature into the feature learning model to represent the data sample and updating, by using the data sample, weights associated with the plurality of features and the feature when the data sample satisfies the first threshold;
determining whether the reconstruction error satisfies a second threshold;
ignoring the data sample when the reconstruction error satisfies the second threshold; and
updating, by using the data sample, weights associated with the plurality of features of the feature learning model when the reconstruction error satisfies neither the first threshold nor the second threshold.