US 11,960,904 B2
Utilizing machine learning models to predict system events based on time series data generated by a system
Femida Eranpurwala, Navi Mumbai (IN); Satyan Kumar, Alpharetta, GA (US); Rahul Maheshwari, Schaumburg, IL (US); and Balaji Poonkundran, Electronic (IN)
Assigned to Accenture Global Solutions Limited, Dublin (IE)
Filed by Accenture Global Solutions Limited, Dublin (IE)
Filed on Jan. 4, 2022, as Appl. No. 17/568,282.
Claims priority of application No. 202141053577 (IN), filed on Nov. 22, 2021.
Prior Publication US 2023/0185579 A1, Jun. 15, 2023
Int. Cl. G06F 15/177 (2006.01); G06F 9/4401 (2018.01); G06F 18/214 (2023.01); G06N 20/00 (2019.01)
CPC G06F 9/4411 (2013.01) [G06F 9/4403 (2013.01); G06F 18/214 (2023.01); G06N 20/00 (2019.01); G06F 9/4401 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, by a device, historic time series data identifying events associated with a system;
performing, by the device, block bootstrapping of the historic time series data, based on a parameter, to generate blocks of data points of the historic time series data,
wherein the parameter is modified based on feedback associated with a prediction;
processing, by the device, the blocks of data points, with first machine learning models, to calculate first predictions;
applying, by the device, first weights to the first predictions to generate weighted first predictions;
aggregating, by the device, the weighted first predictions to generate an aggregated first prediction;
processing, by the device, the blocks of data points, with second machine learning models, to calculate second predictions;
applying, by the device, second weights to the second predictions to generate weighted second predictions;
aggregating, by the device, the weighted second predictions to generate an aggregated second prediction;
processing, by the device, the blocks of data points, with third machine learning models, to calculate third predictions;
applying, by the device, third weights to the third predictions to generate weighted third predictions;
aggregating, by the device, the weighted third predictions to generate an aggregated third prediction;
applying, by the device, final weights to the aggregated first prediction, the aggregated second prediction, and the aggregated third prediction to generate weighted aggregated predictions;
aggregating, by the device, the weighted aggregated predictions to generate a final prediction; and
performing, by the device, one or more actions based on the final prediction.