Real-time ensemble evaluation

A computer program product comprising:

• one or more computer readable storage media having computer-readable program instructions stored on the one or more computer readable storage media, the program instructions executing a computer-implemented method comprising:
• creating a respective coincidence matrix for each component model of a deployed ensemble model;
• generating evaluation measures for each of the component models of the deployed ensemble model, wherein the evaluation measures are based on predictions from respective component models ingesting real-time engine performance data;
• responsive to the respective coincidence matrix failing to exceed a record threshold, updating the respective coincidence matrix with the generated evaluation measures;
• responsive to the respective updated coincidence matrix failing to exceed a disable cut point, flagging the respective component model;
• dynamically adjusting an elimination cut point based on computational constraints associated with the deployed ensemble model;
• and
• responsive to the respective updated coincidence matrix associated with the respective flagged component model not exceeding the elimination cut point, removing the respective flagged component model from the deployed ensemble model.

A computer system comprising:

• one or more computer processors;
• and
• one or more computer readable storage media having computer readable program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more processors, the stored program instructions execute a computer-implemented method comprising:
• creating a respective coincidence matrix for each component model of a deployed ensemble model;
• generating evaluation measures for each of the component models of the deployed ensemble model, wherein the evaluation measures are based on predictions from respective component models ingesting real-time engine performance data;
• responsive to the respective coincidence matrix failing to exceed a record threshold, updating the respective coincidence matrix with the generated evaluation measures;
• responsive to the respective updated coincidence matrix failing to exceed a disable cut point, flagging the respective component model;
• dynamically adjusting an elimination cut point based on computational constraints associated with the deployed ensemble model;
• and
• responsive to the respective updated coincidence matrix associated with the respective flagged component model not exceeding an elimination cut point, removing the respective flagged component model from the deployed ensemble model.