| CPC G06F 9/4494 (2018.02) | 17 Claims |
|
1. An apparatus for determining system model comparison, wherein the apparatus comprises:
at least a processor; and
a memory communicatively linked to the at least a processor, wherein the memory contains instructions configuring the at least a processor to:
receive a first plurality of system data, wherein the first plurality of system data representing a first state of a system, wherein the first plurality of system data comprises first operational and performance metrics;
receive a second plurality of system data representing a second state of the system, wherein the second plurality of system data comprises second operational and performance metrics;
generate a first model of the system, using the first plurality of system data;
generate a second model of the system, using the second plurality of system data;
output a first model output using the first model of the system and the second plurality of system data;
modify the second plurality of system data using at least a perturbation function;
output a second model output using the second model of the system and the modified second plurality of system data applied in the second model; and
train a machine learning model, wherein training the machine learning model comprises:
receiving a training dataset comprising outputs from at least the first model correlated to corresponding evaluations of the at least the first model;
adjusting the machine learning model using the training dataset and a loss function;
comparing new outputs of the adjusted machine learning model to the training dataset;
processing the new outputs of the adjusted machine learning model with the outputs of at least the first model as a function of at least a learned pattern of the adjusted machine learning model; and
generating a score for the outputs of the at least the first model using the adjusted machine learning model, wherein the score for the outputs of the at least the first model indicates a quality value of the outputs of the at least the first model, wherein the quality value is compared against a quality benchmark; and
compare the first model output to the second model output using the trained machine learning model and the generated score;
generate a state change output corresponding to the system utilizing the trained machine learning model and the generated score, wherein the state change output comprises at least a proposed adjustment to a plurality of parameters of the system mirroring the perturbation function applied in the second model; and
present the proposed adjustment as a graphical representation on a user interface including interactive elements by activating event handlers of the user interface to refine system performance of the system based on the proposed adjustment.
|