AbstractInvestment management is a complex, yet promising, domain for the application of intelligent automation. In this paper we discuss the Diversified Investment Management Expert (DIME) system which is under development. DIME is designed to operate in a dynamic environment and hence one of its salient features is its knowledge acquisition via learning. The system starts out with some rules built into it, and machine‐learning techniques are used to acquire further knowledge with time. In this paper we focus on howsimilarity‐based learning(SBL) is used to acquire new rules andexplanation‐based learning(EBL) to refine the built‐in rules and the newly acquired ones.