Demographic-aware federated machine learning

An apparatus comprising a processor and memory including program code, the memory and the program code configured to, with the processor, cause the apparatus to at least:

• receive a plurality of local machine learning model data objects, wherein each local machine learning model data object of the plurality of local machine learning model data objects is (i) received from a respective model data object provider agent of a plurality of model data object provider agents, (ii) associated with a respective inference identifier of a plurality of inference identifiers, (iii) associated with a respective model provider demographic profile of a plurality of model provider demographic profiles comprising demographic data of the respective model data object provider agent is received, wherein the plurality of model data object provider agents is mapped in a multi-dimensional space based at least in part on the plurality of model provider demographic profiles;
• for each selected local machine learning model data object comprising a selected inference identifier and a selected model provider demographic profile, generate a respective global machine learning model data object based at least in part on a plurality of related local machine learning model data objects of the plurality of local machine learning model data objects, wherein (i) each related local machine learning model data object of the plurality of related local machine learning model data objects is associated with the selected inference identifier and the selected model provider demographic profile, (ii) the selected local machine learning model data object is selected based at least in part on a mapping of the multi-dimensional space, and (iii) the selected inference identifier is associated with a per-inference universal machine learning model data object associated with a per-inference statistical relevance measure;
• and
• provide, based at least in part on each of the respective global machine learning model data object, a demographic-aware predictive data analysis application programming interface (API) that is accessible by the plurality of model data object provider agents.

A computer program product comprising at least one non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions configured to:

• receive a plurality of local machine learning model data objects, wherein each local machine learning model data object of the plurality of local machine learning model data objects is (i) received from a respective model data object provider agent of a plurality of model data object provider agents, (ii) associated with a respective inference identifier of a plurality of inference identifiers, (iii) associated with a respective model provider demographic profile of a plurality of model provider demographic profiles comprising demographic data of the respective model data object provider agent is received, wherein the plurality of model data object provider agents is mapped in a multi-dimensional space based at least in part on the plurality of model provider demographic profiles;
• for each selected local machine learning model data object comprising a selected inference identifier and a selected model provider demographic profile, generate a respective global machine learning model data object based at least in part on a plurality of related local machine learning model data objects of the plurality of local machine learning model data objects, wherein (i) each related local machine learning model data object of the plurality of related local machine learning model data objects is associated with the selected inference identifier and the selected model provider demographic profile, (ii) the selected local machine learning model data object is selected based at least in part on a mapping of the multi-dimensional space, and (iii) the selected inference identifier is associated with a per-inference universal machine learning model data object associated with a per-inference statistical relevance measure;
• and
• provide, based at least in part on each of the respective global machine learning model data object, a demographic-aware predictive data analysis application programming interface (API) that is accessible by the plurality of model data object provider agents.