one
SEQUENTIAL MACHINE LEARNING FOR DATA MODIFICATION

Abstract:

In some implementations, a device may receive a request associated with a data record, the request including a first set of input for a first machine learning model, and the first set of input defining a first set of values associated with a first set of parameters that correspond to the data record. The device may obtain, from a data storage device, a second set of values associated with a second set of parameters that correspond to the data record. The device may determine, using the first machine learning model, a score for the data record based on the first set of input and the second set of values, wherein the first machine learning model is trained to receive the first set of input and the second set of values and produce, as output, the score. The device may determine based on determining that the score satisfies a threshold, and using a second machine learning model, a third set of values for the first set of parameters, wherein the second machine learning model is trained, based on historical information associated with the data record, to receive at least a portion of the second set of values as input and produce, as output, the third set of values. The device may modify the data record by updating the first set of parameters using the third set of values.