US 12,032,549 B2
Techniques for creating and utilizing multidimensional embedding spaces
Austin Grant Walters, Savoy, IL (US); Mark Louis Watson, Sedona, AZ (US); Jeremy Edward Goodsitt, Champaign, IL (US); Anh Truong, Champaign, IL (US); and Reza Farivar, Champaign, IL (US)
Assigned to Capital One Services, LLC, McLean, VA (US)
Filed by Capital One Services, LLC, McLean, VA (US)
Filed on Jul. 26, 2022, as Appl. No. 17/873,919.
Application 17/873,919 is a continuation of application No. 16/844,541, filed on Apr. 9, 2020, granted, now 11,429,582.
Prior Publication US 2023/0016044 A1, Jan. 19, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 16/22 (2019.01); G06F 16/28 (2019.01); G06N 20/00 (2019.01)
CPC G06F 16/2264 (2019.01) [G06F 16/2237 (2019.01); G06F 16/285 (2019.01); G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus, comprising:
at least one processor; and
at least one memory storing instructions that, when executed by the at least one processor, cause the at least one processor to:
generate an embedding space customized for one or more data characteristics of a set of data objects, the embedding space customized to the set of data objects via generating the embedding space using a plurality of dimensions defined by one or more embedding space parameters, each of the one or more parameters corresponding to one of the one or more data characteristics, the plurality of dimensions comprising an object depth dimension indicating a number of layers or a number of embedded objects of a corresponding data object, and
generate a set of object vectors for the set of data objects, the set of object vectors configured as a vector of continuous numbers for input to a machine learning classifier or algorithm, the set of object vectors comprising an object vector for each data object in the set of data objects, each object vector in the set of object vectors comprising a depth dimension value indicating the number of layers or the number of embedded objects for the data object, wherein each depth dimension value maps to a corresponding object depth dimension.