US 11,675,877 B2
Method and system for federated deployment of prediction models using data distillation
Paulo Abelha Ferreira, Rio de Janeiro (BR); and Vinicius Michel Gottin, Rio de Janeiro (BR)
Assigned to EMC IP Holding Company LLC, Hopkinton, MA (US)
Filed by EMC IP Holding Company LLC, Hopkinton, MA (US)
Filed on Aug. 31, 2021, as Appl. No. 17/462,509.
Prior Publication US 2023/0068179 A1, Mar. 2, 2023
Int. Cl. G06F 16/00 (2019.01); G06F 18/214 (2023.01); G06F 16/23 (2019.01); G06F 16/28 (2019.01); G06F 18/21 (2023.01)
CPC G06F 18/214 (2023.01) [G06F 16/2358 (2019.01); G06F 16/2365 (2019.01); G06F 16/285 (2019.01); G06F 18/217 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A method for managing data nodes of data node clusters, the method comprising:
obtaining, by a data node manager, a request to deploy a model to a data node;
in response to obtaining the model deployment request:
identifying, by the data node manager, a data node cluster associated with the data node using a data node cluster registry stored in a storage of the data node manager, wherein the data node cluster comprises a plurality of data nodes as specified in the data node cluster registry, wherein the plurality of data nodes comprises the data node as specified in the data node cluster registry;
making a first determination, by the data node manager, that the data node cluster is associated with an available distilled dataset, wherein:
the data node cluster registry further comprises the available distilled dataset associated with the data node cluster, and
the available distilled dataset was generated using distilled dataset update parameters generated by each of the plurality of data nodes of the data node cluster; and
in response to the first determination:
generating, by the data node manager, a model using the available distilled dataset; and
deploying, by the data node manager, the model to the data node.