Techniques for Dynamic Data Validation

原文（中文）

Techniques for dynamic data validation are disclosed herein. An example computer-implemented method includes receiving entity data associated with an entity, the entity data including locations of the entity at respective times. The method further includes determining, by executing a dynamic period algorithm, periods based on the entity data; and applying a machine learning (ML) model to the entity data and the periods. Applying the ML model includes determining, for at least one period, one or more confidence values associated with each location at the respective times included in the period based on (i) a frequency associated with each location and (ii) a period distance value relating a current time to the period. The ML model also outputs a ranking for each location included in the period based on the one or more confidence values. The method further includes generating a data object indicating one or more of the ranked locations.