Methods and apparatus for machine learning predictions and multi-objective optimization of manufacturing processes

A non-transitory computer-readable medium comprising code which, when executed by a processor, cause the processor to:

• train a first regression machine learning model with a first training set to produce a first predictive value for a pending manufacturing request, the first training set including data extracted from a first plurality manufacturing transactions submitted by a requestor entity of a supply chain;
• train a second regression machine learning model with a second training set to produce at least one second predictive value for the pending manufacturing request, the second training set including data from a second plurality manufacturing processes responses received by a set of entities of the supply chain;
• and
• instantiate a multi-objective optimization model that is configured to (1) receive an input including the first predictive value, the second predictive value, a set of features of a physical object, and (2) generate a non-deterministic output with a set of attributes associated with a manufacture process of the physical object, the non-deterministic output complying with a multi-objective condition satisfied in the multi-objective optimization model.

A method, comprising;

• receiving, at a compute device, a manufacturing request that includes an electronic file with a digital model representative of a physical object;
• determining from the digital model a set of features associated with the physical object;
• producing at least one predictive value for the manufacturing request based on at least one regression machine learning model and the set of features associated with the physical object;
• and
• generating a non-deterministic response based on (1) the at least one predictive value for the manufacturing request, and (2) a multi-objective optimization model, the non-deterministic response including a set of attributes associated with a manufacture process of the physical object, the set of attributes complying with a multi-objective condition satisfied in the multi-objective optimization model.