US 11,853,851 B2
Systems and methods for training and employing machine learning models for unique string generation and prediction
Peter Keyngnaert, Landegem (BE); Jan Waerniers, Aalter (BE); Ann Smet, Stekene (BE); and Akanksha Mishra, Boston, MA (US)
Assigned to Camelot UK Bidco Limited, London (GB)
Filed by Camelot UK Bidco Limited, London (GB)
Filed on Oct. 3, 2019, as Appl. No. 16/592,137.
Claims priority of provisional application 62/740,820, filed on Oct. 3, 2018.
Prior Publication US 2020/0111021 A1, Apr. 9, 2020
Int. Cl. G06N 20/00 (2019.01); G06F 40/30 (2020.01); G06N 7/01 (2023.01)
CPC G06N 20/00 (2019.01) [G06F 40/30 (2020.01); G06N 7/01 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A method for generating strings based on one or more seed strings, the method comprising:
training a plurality of machine learning models;
deriving a plurality of random walk models from the trained plurality of machine learning models;
receiving one or more seed strings as an input to the plurality of random walk models;
automatically iteratively generating, by each of the plurality of random walk models, one or more next characters for the one or more seed strings to generate at least one term from each of the random walk models, the at least one term from each of the random walk models includes all characters in the one or more seed strings in addition to the one or more next characters; and
outputting a ranked order for the at least one term generated by each of the random walk models to a graphical user interface.