US 11,853,400 B2
Distributed machine learning engine
Paul Green, Londonderry, NH (US); and Jerzy Bala, Potomac Falls, VA (US)
Assigned to Bottomline Technologies, Inc., Portsmouth, NH (US)
Filed by Bottomline Technologies, Inc., Portsmouth, NH (US)
Filed on Mar. 20, 2023, as Appl. No. 18/123,529.
Application 18/123,529 is a continuation of application No. 17/864,704, filed on Jul. 14, 2022, granted, now 11,609,971, issued on Mar. 21, 2023.
Application 17/864,704 is a continuation of application No. 16/355,985, filed on Mar. 18, 2019, granted, now 11,416,713, issued on Aug. 16, 2022.
Prior Publication US 2023/0244758 A1, Aug. 3, 2023
Int. Cl. G06F 18/24 (2023.01); G06N 20/00 (2019.01); G06F 18/40 (2023.01); G06F 18/214 (2023.01); G06F 16/28 (2019.01); G06N 5/025 (2023.01)
CPC G06F 18/24765 (2023.01) [G06F 16/285 (2019.01); G06F 18/2148 (2023.01); G06F 18/40 (2023.01); G06N 5/025 (2013.01); G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A machine learning apparatus comprising:
a first data storage device including a first distributed data set;
a first network connector connected to a network, the first network connector in communications with a second network connector on a second data storage device on a machine learning server, the second data storage device including a second distributed data set;
a model orchestrator, stored in the first data storage device and executing on the machine learning apparatus, the model orchestrator programmed to publish a set of data identifiers including data elements and data features, and programmed to send the set of the data identifiers through the first network connector to the second network connector to a second prediction manager executing on the machine learning server;
a first prediction manager connected to the first data storage device programmed to receive the set of the data identifiers from the model orchestrator and to calculate a first quality control metric and a first rule set using a first machine learning algorithm on the first distributed data set;
a prediction orchestrator programmed to receive the first quality control metric and the first rule set from the first prediction manager and to receive from the second prediction manager a second quality control metric and a second rule set determined from the second distributed data set; and
the prediction orchestrator further programmed to combine the first rule set and the second rule set into a common rule set and to combine the first quality control metric and the second quality control metric into a combined quality control metric.