US 12,288,009 B2
Apparatus and method for option data object performance prediction and modeling
Leopold Silberstein, Chicago, IL (US); Addhyan Pandey, Chicago, IL (US); Abhaya Parthy, Palo Alto, CA (US); and Boris Lerner, Chicago, IL (US)
Assigned to BYTEDANCE INC., Wilmington, DE (US)
Filed by Bytedance Inc., Wilmington, DE (US)
Filed on Jun. 22, 2021, as Appl. No. 17/354,549.
Application 17/354,549 is a continuation of application No. 15/840,364, filed on Dec. 13, 2017, granted, now 11,087,045.
Claims priority of provisional application 62/434,163, filed on Dec. 14, 2016.
Prior Publication US 2021/0357553 A1, Nov. 18, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 30/20 (2020.01); G06F 111/02 (2020.01); G06F 111/10 (2020.01)
CPC G06F 30/20 (2020.01) [G06F 2111/02 (2020.01); G06F 2111/10 (2020.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus comprising at least one processor and at least one memory coupled to the at least one processor, wherein the at least one processor is configured to at least:
receive an option data object;
extract from the option data object a set of option data object parameters comprising a set of categorical parameters comprising information associated with a particular category of the option data object, and a set of non-categorical parameters;
input, to a first predictive model, the set of categorical parameters;
receive, from the first predictive model, a first weighted value associated with the option data object, wherein the first weighted value comprises a first predicted performance level of the option data object generated based at least in part on the set of categorical parameters;
input, to a second predictive model, the first weighted value and the set of non- categorical parameters, wherein the second predictive model comprises a random forest and is different from the first predictive model;
receive, from the second predictive model a second weighted value associated with the option data object, wherein the second weighted value comprises a second predicted performance level of the option data object;
generate a modified data record by modifying a data record associated with the option data object to include the second weighted value, wherein the modified data record comprises an indication if the option data object has not previously been presented; and
based at least in part on the modified data record and the indication, cause a reallocation of a set of resources associated with the option data object.