US 11,755,954 B2
Scheduled federated learning for enhanced search
Ali Anwar, San Jose, CA (US); Syed Amer Zawad, Reno, NV (US); Yi Zhou, San Jose, CA (US); and Nathalie Baracaldo Angel, San Jose, CA (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Mar. 11, 2021, as Appl. No. 17/199,403.
Prior Publication US 2022/0292392 A1, Sep. 15, 2022
Int. Cl. G06F 16/00 (2019.01); G06N 20/00 (2019.01); G06F 16/2453 (2019.01)
CPC G06N 20/00 (2019.01) [G06F 16/2453 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
maintaining an indication of availability over time and resource usage for each computing device of a plurality of computing devices;
determining an optimal combination of a subset of the plurality of computing devices for each round of one or more rounds of training based on the availability over time and the resource usage for each computing device;
generating a global model utilizing the one or more optimal combinations of the plurality of computing devices; and
performing a query utilizing the global model.