MODEL OPTIMIZATION AND STABILIZATION USING QUANTUM COMPUTING

Aspects of the disclosure relate to machine learning and quantum computing. A computing platform may receive historical information, which may include feature information and rate of change information. The computing platform may train, a ML model, by inputting the feature information and the rate of change information, which may make the ML model stable against data drift. The computing platform may receive a first query, input the first query into the ML model, to identify a solution to the first query, which may include identifying a first solution, perturbing the first solution a number of times, and ultimately identifying a second solution, more accurate than the first solution. The computing platform may send this second solution to the user device.