US 12,106,407 B2
Systems and methods for generating a single-index model tree
Agus Sudjianto, Charlotte, NC (US); Aijun Zhang, Island E (HK); and Zebin Yang, Island E (HK)
Assigned to Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed by Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed on May 2, 2023, as Appl. No. 18/311,046.
Application 18/311,046 is a continuation of application No. 17/305,352, filed on Jul. 6, 2021, granted, now 11,688,113.
Prior Publication US 2023/0267661 A1, Aug. 24, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 11/20 (2006.01); G06F 30/20 (2020.01); G06N 20/20 (2019.01); G06Q 10/067 (2023.01); G06Q 30/0201 (2023.01); G06Q 30/0204 (2023.01); G06Q 40/03 (2023.01)
CPC G06T 11/206 (2013.01) [G06F 30/20 (2020.01); G06N 20/20 (2019.01); G06Q 10/067 (2013.01); G06Q 30/0205 (2013.01); G06Q 30/0206 (2013.01); G06Q 40/03 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A method for generating a single-index model (SIM) tree, the method comprising:
while a maximum tree depth has not been reached, iteratively:
(i) generating a SIM estimation for one or more nodes of a tree level based on a data set,
(ii) for each node of the one or more nodes, selecting a split point and a split variable based on the SIM estimation for that node, and
(iii) based on the selected split points and split variables for the nodes at the tree level, generating a set of nodes for a next tree level; and
in response to generation of nodes for the maximum tree depth, fitting a SIM for each leaf node of the SIM tree based on a subset of the data set represented by that leaf node.