US 11,868,166 B2
Repairing machine learning pipelines
Julian Timothy Dolby, Bronx, NY (US); Jason Tsay, White Plains, NY (US); and Martin Hirzel, Ossining, NY (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Aug. 5, 2021, as Appl. No. 17/444,498.
Prior Publication US 2023/0059857 A1, Feb. 23, 2023
Int. Cl. G06F 11/36 (2006.01); G06N 20/00 (2019.01)
CPC G06F 11/3692 (2013.01) [G06F 11/3684 (2013.01); G06F 11/3688 (2013.01); G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for detecting and correcting errors in one or more machine learning pipelines, the computer-implemented method comprising:
generating, by a computing device, a plurality of test machine learning pipeline instances based upon a target machine learning pipeline;
evaluating, by the computing device, the plurality of test machine learning pipeline instances for failure in a task;
identifying, by the computing device, one or more root causes of error based upon the evaluated plurality of test machine learning pipeline instances and failure in the task;
creating, by the computing device, a remediated target machine learning pipeline based upon the identified one or more root causes of error; and
outputting the remediated machine learning pipeline that repairs the one or more root causes of error in the plurality of test machine learning pipeline instances, wherein the remediated machine learning pipeline depicts and details differences between the remediated machine learning pipeline and the plurality of test machine learning pipeline instances, the identified one or more root causes of corruption, and where the results were implemented in the plurality of test machine learning pipeline instances.