US 12,242,940 B1
Dynamic simulation analytics
John Clare Brocklebank, Franklin, TN (US); Ann L. Cutrell, Durham, NC (US); Savera Tanwir, Morrisville, NC (US); and William Cyrus Bradford, Durham, NC (US)
Assigned to SAS INSTITUTE INC., Cary, NC (US)
Filed by SAS Institute Inc., Cary, NC (US)
Filed on Jun. 4, 2024, as Appl. No. 18/733,296.
Claims priority of provisional application 63/556,704, filed on Feb. 22, 2024.
Claims priority of provisional application 63/618,755, filed on Jan. 8, 2024.
Claims priority of provisional application 63/606,460, filed on Dec. 5, 2023.
Int. Cl. G06N 20/00 (2019.01)
CPC G06N 20/00 (2019.01) 30 Claims
OG exemplary drawing
 
1. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, the computer-program product including instructions operable to cause a computing system to:
display, at a graphical user interface, a selection element for selecting a simulated system;
receive, using the graphical user interface, a selection of the simulated system;
responsive to the selection of the simulated system, obtain, by the computing system, an initial predicted state for the simulated system according to an initial set of inputs to a computer model, wherein the computer model is generated based on data of at least two data types that are different data types;
display, at the graphical user interface, a control element for controlling a predicted output for the simulated system;
receive, at the graphical user interface, a manipulation of the control element, wherein the manipulation indicates to the computing system a request for a computer-generated set of inputs that will generate, for the simulated system, a user-requested change in the initial predicted state according to the manipulation of the control element;
responsive to the manipulation of the control element, the computing system:
generates a set of decision deltas for the computer model,
wherein each decision delta of the set of decision deltas corresponds to a respective one of the at least two data types,
wherein a first decision delta corresponding to a first data type of the at least two data types indicates an extent of allowable change of a first initial input of the initial set of inputs; and
wherein a second decision delta corresponding to a second data type of the at least two data types indicates an extent of allowable change of a second initial input of the initial set of inputs;
obtains a computer function for the set of decision deltas;
determines allowable function inputs to the computer function, wherein the allowable function inputs are derived based on the set of decision deltas, the user-requested change, and the computer model; and
computes, using one or more of the allowable function inputs, at least one minimum or maximum value for the computer function; and
based on the at least one minimum or maximum value for the computer function, output, to the graphical user interface, computer-generated output information based on the computer-generated set of inputs that, according to the computer model, will affect the user-requested change in the initial predicted state in the simulated system, wherein the computer-generated set of inputs comprise:
a first computer-generated input corresponding to the first data type that is different from the first initial input; and
a second computer-generated input corresponding to the second data type that is different from the second initial input.