Systems, methods, and apparatuses for predicting resource shortages using machine learning

A computer program product for predicting resource shortages using machine learning, wherein the computer program product is structured for dynamic medical product shortage preclusion and remediation based on determining a future supply shortage for one or more resource identifiers for healthcare products, wherein the computer program product comprises non-transitory computer-readable program code portions embodied therein, the non-transitory computer-readable program code portions which when executed by a processing device are configured to cause the processor to:

• identify at least one resource identifier from a resource database as a candidate for shortage prediction, wherein the at least one resource identifier corresponds to a healthcare product resource comprising at least one of a physical item, a product, a device, or an item of manufacture;
• receive a resource transaction dataset associated with the at least one resource identifier, wherein the resource transaction dataset comprises resource transaction data parameters associated with the resource comprising at least a source or manufacturer of the resource, a critical level of resource, and substitutions for the resource;
• identify at least one key performance indicator from the resource transaction data parameters of the resource transaction dataset that is associated with predicting a resource supply shortage of a resource;
• generate, based on the at least one key performance indicator, at least one key performance signal, wherein the at least one key performance signal comprises a plurality of values of the at least one key performance indicator over a predetermined period;
• apply the at least one key performance signal to a supply shortage labeling operation, comprising identifying patterns within the at least one key performance signal that are associated with the resource supply shortage of the resource;
• identify, by the supply shortage labeling operation, at least one resource shortage event associated with the at least one resource identifier;
• apply the at least one key performance signal and the at least one resource shortage event to train a supply shortage machine learning model;
• and
• generate, by the supply shortage machine learning model, a likelihood of a future-shortage for the resource identifier, wherein the likelihood of a future-shortage for the resource identifier is based on the at least one key performance signal for at least one point in time.

A computer implemented method for predicting resource shortages using machine learning, wherein the computer implemented method is structured for dynamic medical product shortage preclusion and remediation based on determining a future supply shortage for one or more resource identifiers for healthcare products, the computer implemented method comprising:

• identifying at least one resource identifier from a resource database as a candidate for shortage prediction, wherein the at least one resource identifier corresponds to a healthcare product resource comprising at least one of a physical item, a product, a device, or an item of manufacture;
• receiving a resource transaction dataset associated with the at least one resource identifier, wherein the resource transaction dataset comprises resource transaction data parameters associated with the resource comprising at least a source or manufacturer of the resource, a critical level of resource, and substitutions for the resource;
• identifying at least one key performance indicator from the resource transaction data parameters of the resource transaction dataset that is associated with predicting a resource supply shortage of a resource;
• generating, based on the at least one key performance indicator, at least one key performance signal, wherein the at least one key performance signal comprises a plurality of values of the at least one key performance indicator over a predetermined period;
• applying the at least one key performance signal to a supply shortage labeling operation, comprising identifying patterns within the at least one key performance signal that are associated with the resource supply shortage of the resource;
• identifying, by the supply shortage labeling operation, at least one resource shortage event associated with the at least one resource identifier;
• applying the at least one key performance signal and the at least one resource shortage event to train a supply shortage machine learning model;
• and
• generating, by the supply shortage machine learning model, a likelihood of a future-shortage for the resource identifier, wherein the likelihood of a future-shortage for the resource identifier is based on the at least one key performance signal for at least one point in time.