Advancing Dynamic Evolutionary Optimization Using In-Memory Database Technology

This paper reports on IMDEA (In-Memory database Dynamic Evolutionary Algorithm), an approach to dynamic evolutionary optimization exploiting in-memory database (IMDB) technology to expedite the search process subject to change events arising at runtime. The implemented system benefits from optimization knowledge persisted on an IMDB serving as associative memory to better guide the optimizer through changing environments. For this, specific strategies for knowledge processing, extraction and injection are developed and evaluated. Moreover, prediction methods are embedded and empirical studies outline to which extent these methods are able to anticipate forthcoming dynamic change events by evaluating historical records of previous changes and other optimization knowledge managed by the IMDB.