| CPC G06Q 10/06395 (2013.01) [G06N 20/00 (2019.01); G06Q 50/04 (2013.01); G06T 7/0004 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30108 (2013.01)] | 16 Claims |
|
1. A method for making predictions relating to products manufactured via a manufacturing process, the method comprising:
receiving, by a processor, a plurality of input vectors associated with a plurality of output values and a plurality of time intervals;
clustering, by the processor, a first set of input vectors of the plurality of input vectors into a first time interval of the plurality of time intervals, and a second set of input vectors of the plurality of input vectors into a second time interval of the plurality of time intervals;
labeling the first time interval with a first label and the second time interval with a second label;
training, by the processor, a first machine learning model for a first time interval of the plurality of time intervals, and a second machine learning model for a second time interval of the plurality of time intervals, the training of the first machine learning model being based on the first set of input vectors associated with the first time interval and first ones of the output values associated with the first set of input vectors, and the training of the second machine learning model being based on the second set of input vectors associated with the second time interval and second ones of the output values associated with the second set of input vectors;
training, by the processor, a classifier machine learning model separate from the first machine learning model for selecting a time interval of the plurality of time intervals, wherein the training is based on the first set of input vectors, the second set of input vectors, the first label, and the second label;
executing, by the processor, the classifier machine learning model based on input data received for a product, and receiving as output of the classifier machine learning model the first label of the first time interval into which the input data for the product is classified;
selecting based on the first label the first machine learning model from among the first machine learning model associated with the first time interval and the second machine learning model associated with the second time interval; and
executing, by the processor, the first machine learning model for predicting an output based on the input data.
|