Skip to Main Content
This paper discusses how to use knowledge in machine learning, so that decision-making with knowledge-based machine learning can be obtained. Objectives, anti-objectives and it is support of decisions are formally defined, and the knowledge base, evidence base, sample base and qualified samples for decision problem are discussed in details. This paper discusses the uncertainty in the dicision problem with a method of how to use knowledge for measuring the supports of our objectives form knowledge in our hands. Three functions, namely: the initial evaluation function, sample evaluation function, making the evaluation function are proposed.