Abstract: There is a need for more effectively and efficiently performing predictive data analysis to determine associations between input data objects and predictive categories This need can be addressed by, for example, solutions for performing predictive data analysis to determine associations between input data objects and predictive categories that utilize at least one of accuracy scores for predictive categories, evidentiary scores for predictive categories, and predicted certification statuses for claim data objects. In one example, a method includes: for each predictive category of one or more predictive categories associated with a claim data object, determining an accuracy score and an evidentiary score; determining a predicted certification status for the claim data object based on each accuracy score for a predictive category and each evidentiary score for a predictive category; and performing prediction-based actions based on each predicted certification status.