System and method for identifying and quantifying deviations and inconsistent features in machine-learning models

A method, comprising:

• accessing one or more model performance metrics associated with a machine-learning model, wherein the machine-learning model is identified as having generated one or more prediction outputs that is deviating from an expected prediction output;
• executing a hybrid machine-learning model trained to identify one or more patterns of deviation of the one or more prediction outputs and one or more inconsistent features in a data set associated with the machine-learning model based at least in part on the one or more model performance metrics;
• and
• identifying an underperformance of the machine-learning model based at least in part on the identified one or more patterns of deviation and the one or more inconsistent features, wherein identifying the underperformance of the machine-learning model comprises generating an output identifying significant contributors to the machine-learning model having generated the one or more prediction outputs that is deviating from the expected prediction output.

A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to:

• access one or more model performance metrics associated with a machine-learning model, wherein the machine-learning model is identified as having generated one or more prediction outputs that is deviating from an expected prediction output;
• execute a hybrid machine-learning model trained to identify one or more patterns of deviation of the one or more prediction outputs and one or more inconsistent features in a data set associated with the machine-learning model based at least in part on the one or more model performance metrics;
• and
• identify an underperformance of the machine-learning model based at least in part on the identified one or more patterns of deviation and the one or more inconsistent features, wherein identifying the underperformance of the machine-learning model comprises generating an output identifying significant contributors to the machine-learning model having generated the one or more prediction outputs that is deviating from the expected prediction output.