US 12,223,521 B1
Determining a topology for distribution of services using demand aggregation
Julian Enrique Pachon, Austin, TX (US); Marc Joseph Anderson, Driftwood, TX (US); Bhargav Kunkulagunta, Hyderabad (IN); Nader M. Kabbani, Bellevue, WA (US); and Karthik Charan Konduri, Redmond, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Sep. 29, 2021, as Appl. No. 17/489,113.
Claims priority of provisional application 63/222,337, filed on Jul. 15, 2021.
Int. Cl. G06Q 30/0204 (2023.01); G06N 20/00 (2019.01); G06Q 10/0631 (2023.01); G06Q 10/08 (2024.01); G06Q 30/0201 (2023.01)
CPC G06Q 30/0205 (2013.01) [G06N 20/00 (2019.01); G06Q 10/0631 (2013.01); G06Q 10/08 (2013.01); G06Q 30/0206 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
generating, by one or more processors, a first plurality of training data that includes a plurality of information relating to at least one of locations, communities, event types, or trends for provisioning of a service;
training, using the first plurality of training data by the one or more processors, a machine learning model to determine a demand estimation model that specifies a plurality of locations and a plurality of predicted demands for each of the plurality of locations;
generating, by the one or more processors, a user interface configured to obtain a plurality of demand information and a plurality of parameters in connection with providing the service to a population distributed over a geographic area;
causing, by the one or more processors, the user interface to be presented on a client device;
receiving, by the one or more processors and via an interaction with the user interface, the plurality of demand information and the plurality of parameters;
processing, using the trained machine learning model, the plurality of demand information and the plurality of parameters to generate a predicted demand estimation model;
determining, by the one or more processors and based at least in part on the predicted demand estimation model, a plurality of locations at which to provide the service to the population based at least in part on a rate of change of consumption of the service relative to a threshold value;
determining, by the one or more processors and based at least in part on the predicted demand estimation model, an estimated demand forecast for at least one location of the plurality of locations;
determining, by the one or more processors and based at least in part on the plurality of parameters, a cost function representing a total cost for providing the service to the estimated demand forecast associated with the at least one location;
optimizing, by the one or more processors, the cost function to determine an optimized cost function presenting a lowest relative total cost for providing the service to the estimated demand forecast associated with the at least one location;
generating, by the one or more processors, a topology associated with the optimized cost function;
generating, by the one or more processors, a second user interface presenting a representation of the topology, wherein the representation of the topology specifies a plurality of topology information associated with the topology and the plurality of topology information includes at least one of a capacity associated with the at least one location, a facility type associated with the at least one location, or an event type associated with the at least one location;
causing, by the one or more processors, the second user interface to be presented on the client device;
comparing, by the one or more processors, an actual demand associated with the at least one location to the estimated demand forecast;
generating, by the one or more processors, a second plurality of training data based at least in part on the comparison of the actual demand associated with the at least one location to the estimated demand forecast; and
updating, by the one or more processors, the trained machine learning model using the second plurality of training data.